Blockchain Technology Adoption in Saudi Arabia’s Higher

Blockchain Technology Adoption in Saudi Arabia’s

Higher

Education Sector

1 Introduction

1.1 Chapter Summary

This chapter provides an overview of the research background and context as well as its

importance and contribution to the existing knowledge. Section 1.2 offers the study

background: it discusses the importance of the educational sector for national

development and the role of technological innovations in this process. Specific

references are made to the role of blockchain as an emergent technology for education.

Section 1.3 reviews the research context – the higher education sector in Saudi Arabia.

Section 1.4 provides a brief review of blockchain technology and its potential for

education is provided. Section 1.5 identifies the research significance and contribution.

In Section 1.6, the research outline is presented and visualised.

1.2 Background

The educational sector plays an important role in any country’s development. This is

particularly true for countries which are looking for economic diversification and

tapping into their enormous human potential. For Saudi Arabia, the development of

education, in particular higher education, has become one of the major goals of the

consequent Development Plans starting from the 1970s (Allahmorad & Zreik, 2020).

The primary national development document, Vision 2030, assigns higher education one

of the key roles in turning Saudi Arabia to a major global force within the upcoming

decade.

Historically, the developments in the educational sector have been increasingly seen

alongside technological progress (Bernacki, Greene, & Crompton, 2020; Ratheeswari,

2018). Indeed, technology is often seen as a disruptive force in all aspects of education:

from the new approaches in providing learning content to administration and control

over education processes (Selwyn, 2012). In different periods of time, both researchers

and practitioners have examined the impact on education by at-the-time novel

technologies such as mobile networks, cloud computing, virtual reality, and the Internet-

of-Things (AlEmran, Malik, & Al-Kabi, 2020; Ercan, 2010; Eschenbrenner & Nah,

2007; Hedburg & Alexander, 1994). Both the transformative impact and practical

applications of these technologies for education purposes are well reported in literature.

Given the importance of integrating novel technologies into higher education,

administrators and IT specialists in colleges and universities are likely to explore the

opportunities and challenges in applying these technologies within their organizations.

Many different factors may be in play in this case including, but not limited to, the

technology itself and its potential to improve organisational processes; the

organisational structures and existing operational frameworks; the human resources that

the organization possesses; and the environment in which the organization operates. It is

important to understand the influence of these factors to understand both the process of

new technology adoption and its potential to benefit the educational organization.

One of the emerging technologies with a high potential for education today is

blockchain. Introduced primarily as a decentralised cryptocurrency tool (Nakamoto,

2008), blockchain received widespread attention since Buterin’s (2014) ground-breaking

paper which described its many possible applications in different areas of life. Within

only a few years, blockchain applications found their way into various industries,

although practical implementations in many of them remain in their infancy (Grover,

Kar, & Janssen, 2019). The higher education sector is not an exception in this case.

Whereas researchers identified numerous potential uses of blockchain to contribute to

educational institutions’ value chains (e.g., (Alammary, Alhazmi, Almasri, & Gillani,

2019; Awaji, Solaiman, & Albshri, 2020), it is far too early to speak of the mass

adoption of this technology.

Blockchain in Saudi Arabia’s higher education sector remains mostly an unexplored

area. Saudi researchers have so far concentrated on the reviews of blockchain

applications in higher education (Alam & Benaida, 2020; Alammary, Alhazmi, Almasri,

& Gillani, 2019; Malibari, 2020) whereas empirical studies are virtually absent. To fill

this gap and to offer practical guidance for blockchain implementation in Saudi colleges

and universities, this study develops and tests a framework for blockchain adoption in

the Saudi higher education sector. Specifically, the proposed model investigates the

influence of organisational, technological, environmental, and human factors on

blockchain adoption in the context of Saudi higher education institutions. It is expected

that the study results will aid administrators and IT professionals in developing

actionable, practical steps for blockchain integration in educational institutions to

extract the maximum benefit.

1.3 Research Context

Saudi Arabia is the largest country on the Arabian Peninsula with a total area of

2,250,000 square kilometres and an estimated population of about 34.2 million (CIA,

2022). The contemporary education system in Saudi Arabia has its roots in the 1970s.

Prior to this, the traditional kuttab system consisting of religious schools offered limited

education opportunities for privileged families. The absence of a reliable network of

educational institutions available to the masses caused extremely high illiteracy rates:

85% among men and 98% among women by the end the 1960s (Hosen, 2018). The

rapid expansion of educational programs was largely driven by the growing oil revenues

which allowed the country to invest in educational institutions, infrastructure, and

human resources. In 1970, the first Five Year Development Plan emphasized for the first

time the need for education improvements at the national level and outlined the initial

steps in creating a network of education institutions across the country. Within the next

fifteen years, enrolments in elementary schools increased by over 190%, in intermediate

schools by 375%, and in secondary schools by over 700% (Allahmorad & Zreik, 2020).

The government’s commitment to expanding education has remained strong over the

past decades. Saudi Arabia has consistently invested a substantial portion of state

revenue in the educational sector, even in times of falling oil revenue. Experts noted that

this is prompted in part by the realization of the finite hydrocarbon revenue streams in

the future and the need to diversify the economy as a result (Horschig, 2016; Moshashai,

Leber, & Savage, 2020). Starting from the 1990s, spending on education remained one

of the largest budgetary lines ranging from 4% to over 8% share of GDP, one of the

highest among the OECD countries (Euchi, Omri, & Al-Tit, 2018). Today, Saudi Arabia

ranks clearly first among the Gulf Cooperation Council (GCC) countries in terms of

budgeted government education expenditures (Figure 1).

Figure 1: Comparative Expenditures on Education among the GCC Countries as a

Percentage of State Budget, 2021 (Statista, 2021)

Given the importance assigned to education in Saudi Arabia as well as the amount of

government expenditure on it, it is not surprising that the education system in the

country remains administered in a centralized way. The Ministry of Education (MOE) is

the main governing body that formulates education policies and exercises oversight.

Additionally, the Ministry of Higher Education (MHE) and the Technical and Vocational

Training Corporation (TVTC) are controlling government bodies in the corresponding

educational areas and are accountable before the MOE. At regional, municipal and local

levels, the educational policies are implemented and overseen by a large number of

educational departments, directorates, and offices. Some attempts to reduce such

bureaucratic load have been recently undertaken. Specifically, since 2018, about 2,000

educational institutions were granted fiscal and administrative autonomy by the MOE as

well as the ability to implement school curricula changes (Allahmorad & Zreik, 2020).

There are still no available data on the results of this experiment, although it generally

indicates the willingness of the government to produce a more independent education

system similar to those established in developed Western nations.

1.3.1 Higher Education in Saudi Arabia

Higher education has been given special attention in Saudi Arabia within the national

development strategy, especially given the country’s relatively short history of higher

education. The first full scale higher education institution, King Saud University, was

established only in 1957. However, under the educational programs guided by the first

development plans, Saudi Arabia opened six more universities and increased higher

education enrolments from 1969 to over 6900 in the following twenty years (Saleh M. ,

1986). Still, higher education in the country remained more of a privilege than a

universal opportunity up until the 2000s when the focus shifted towards making higher

education accessible to wider social groups. From a comparative perspective, the higher

education gross enrolment ratio in 2000 was about 22%, while in 2018 it reached a huge

68%, which is on par with France and Canada (Figure 2).

Figure 2: Saudi Arabia’s Gross Enrolment Rate for Higher Education: 1971-2018

(Allahmorad & Zreik, 2020)

This growth in enrolments corresponded with the growth in higher education offerings

both inside and outside of the Kingdom. On the home front, the number of public

universities expanded to 29 by 2020, with additional hundreds of colleges and

vocational training schools available. However, public institutions alone have been

unable to satisfy the growing higher education demand. As a result, the MHE allowed

the development of the private higher education sector in an attempt to offer additional

educational capacity, reduce public education costs and introduce more options for

higher education attainment. In the beginning of 2021, the MOE accredited 14 private

universities and about 40 college programs (Saudi Ministry of Education, 2022). While

the total enrolment in private higher education institutions is about 5% of the total

enrolment, it is expected that they will play a growing role in the education sector

development by absorbing excess capacity of the quickly growing college-age

population in Saudi Arabia (Allahmorad & Zreik, 2020).

In addition to the domestic higher education offerings, Saudi Arabia has become one of

the largest providers of funds in the world for national students seeking education

abroad. The King Abdullah Scholarship Program launched in 2005 introduced full

scholarships for Saudi undergraduate and graduate degree seekers, covering tuition and

living expenses for students and their spouses in more than 30 countries. The program

has produced over 200,000 graduates with foreign university diplomas since its launch

(Kottasova, 2016). The recent global oil price crash prompted some cuts in the program

and although it continues today, it is limited to the top 200 international schools and 50

programs (Allahmorad & Zreik, 2020).

The latest available data at the time of writing showed 1.62 million Saudi students

enrolled in higher education institutions with 90% enrolled in public colleges and

universities and the rest equally distributed between those studying in private schools

and abroad (Allahmorad & Zreik, 2020). An important achievement of the higher

education system is providing opportunities for female students, which was unthinkable

a few decades ago. Figure 3 offers an outlook of the student distribution by gender

across six higher education programs. It can be seen that female students are equally

represented in Bachelor’s, Higher Diploma 1 , and Master’s programs, although they still

lack representation in Associate, Graduate Medical, and Doctoral programs. In 2009, the

King Abdullah University of Science and Technology (KAUST) became the first

coeducational university, although higher education in the country remains mostly

separated by gender. Another notable change in higher education governance has been

the introduction of King Saud University stakeholders into the decision-making process

and policy formulation in relation to this oldest and largest university in the country.

These moves demonstrate that higher education in Saudi Arabia may be both a driver for

development and experimental ground for relaxing government’s hold on various

aspects of life in Saudi Arabia.

1 Higher Diploma degree in Saudi Arabia is given upon completion of additional coursework after

Bachelor’s degree. The difference between it and a Master’s degree is that such programs are usually

shorter (1 year approximately) and they do not require the completion of a research project.

Figure 3: Distribution of Saudi Students by Number and Gender across the Higher

Education Programs (Allahmorad & Zreik, 2020).

The present development initiatives in the Saudi higher education sector are driven by

two factors: the national development outlook formulated within the Vision 2030 plan

and the integration of education and technology. Vision 2030 is the strategic

development framework for Saudi Arabia introduced in 2016. With the main emphasis

on reducing the dependence on oil, the plan assigns one of the major roles in this

process to education. A description of one of the three major themes of Vision 2030

includes the following:

“a thriving economy provides opportunities for all by building an

education system aligned with market needs and creating

economic opportunities for the entrepreneur, the small enterprise

as well as the large corporation. (Vision 2030, p. 13)

Section 2.1.1 of Vision 2030 also provides the specifics on how education will

contribute to the country’s economic growth. The major goals in this regard are:

- achieving student results in global education indicators above the international

averages;

- placing five Saudi universities in the top global 200 rank by improving the

quality of education;

- developing a reliable network of career counsellors;

- investing in strategic private-public partnerships;

- improving monitoring and statistics on student population dynamics and chosen

majors to close the gap between education outcomes and the market needs.

Technology is seen as an important driver for many of these initiatives and the

development of the higher education sector in general. Looking at the history of

information and communication technology (ICT) integration in the education sector in

Saudi Arabia, three phases can be identified. The first phase took place during the Fifths

and Sixth Development Plans in 1990-2000. It focused on the introduction of ICT

courses to the curriculum and establishing the first IT programs. The second phase took

place during the 7th, 8th, and 9th Plans (2000-2014) and was characterized by large-scale

training programs in ICT for both educators and students as well as the mass

introduction of ICT into education processes. At present, the phase of digital

transformation is taking place. It can be viewed that the phase started in 2016 with the

establishment of the Tatweer Educational Technologies Company by the MOE.

Tatweer, which is essentially a technology arm of the Saudi MOE, is tasked with

developing high-tech solutions and e-services for the education sector. It is working

closely with both private and non-for-profit organizations to achieve the major goals of

the digital transformation of the education sector (Tatweer, 2022). According to the

official Tatweer strategy outlined during the Sustainable Education Meeting 2018, the

process of digital transformation involves: 1) the creation of common digital platforms

across education entities; 2) launching related digital initiatives; and 3) investing in

digital education asset developments (TETCO, 2018). A number of core initiatives have

been launched to achieve this, ranging from upgrading skills of university instructors to

raising education quality and outcomes through digital technologies. However, the most

ambitious and technology-driven initiative is the transfer to smart schools.

According to the Tatweer site (Tatweer, 2022), the company sees smart schools as

completely transforming the traditional learning experience. This goes beyond simply

introducing smart technologies for classrooms; rather, it can be seen as a new approach

to learning mediated by modern technologies. Some specific aspects of smart schools

are outlined as follows:

1) Interactive digital learning environment via common virtual platforms to connect

anyone, anywhere;

2) Utilization of applications and platforms on the bases of those commonly used

by learners for more intuitive access to educational resources;

3) Integrated learning systems for interactive distance learning;

4) Digitization of common tasks and reducing environmental impact through the

elimination of excess paperwork, ink, and other unnecessary physical attributes;

5) Expanding learning services both inside and outside school boundaries.

While relatively new for Saudi Arabia, the concept of a smart university has been

prominent in most developed nations which underwent the digital revolution earlier

(Uskov, Bakken, Howlett, & Jain, 2018). Recently, a growing body of research emerged

on the incorporation of blockchain technology for smart school initiatives with some

promising applications (Alam & Benaida, 2020; Chen, Xu, Lu, & Chen, 2018; Lam &

Dongol, 2020). In Saudi Arabia, blockchain has been recently deployed by the Saudi

Arabia Monetary Agency (SAMA) for money transfer and deposit security purposes

(Hafiz, 2020). In higher education, however, blockchain applications remain rather

limited. The only notable example has been the KAUST experiments with Blockcerts –

blockchain-based diplomas for its graduates (Rogers C. , 2018). The range of possible

applications of blockchain is much wider, and it can substantially enhance the transfer

towards digital education pursued by the Saudi MHE. The following section reviews

blockchain technology, its applications in higher education and possible barriers to

adoption in Saudi Arabia.

1.4 Blockchain Technology

Blockchain was first described in the ground-breaking paper by Nakamoto in 2008 as a

peer-to-peer distributed ledger to register Bitcoin cryptocurrency transactions

(Nakamoto, 2008). True to its name, blockchain is, in essence, a continuously expanding

chain of records (blocks) linked by cryptography. A sample Bitcoin blockchain is

presented in Figure 4, showing each block contains four elements:

1) an encrypted previous block’s SHA-256 hash: one-way mathematical

algorithm linking data;

2) a trusted timestamp securing time for block creation and/or alteration;

3) a Merkle-tree transaction data;

4) a nonce: an arbitrary string that can only be used once.

Figure 4: Illustration of Blockchain Technology Used for Bitcoins (Wander, 2013)

Blocks are created and checked by all participants in a network. Each node has a replica

of the transaction ledger where it can transmit transactions to the other nodes, check the

ledger against the other nodes, and insert new entries to the ledger when approved by all

nodes (Nakamoto, 2008).

A sample transaction using blockchain technology is the following (as described by

Chen et al., 2018). Node A initiates a transaction with Node B. A cryptographic

combination of public and private keys is used by the network to uniquely identify the

nodes. The transaction is sent to the network memory pool for verification and

validation. When a certain number of approvals from the network nodes is achieved, it

is described as reaching consensus, basically verifying and assigning validity to the

transaction. This is achieved through mining – use of a consensus algorithm to achieve

an updated state of the distribution ledger (Kraft, 2016). A new block is then formed on

the network and updated by every node on their respective ledgers. The block receives a

record of all the transactions which have taken place, a timestamp, and becomes linked

to the previous block with a cryptosignature. Such a process ensures that each

transaction is unique: any attempt to interfere with it or alter it would inevitably break

the existing chain (Chen et al., 2018).

Blockchain technology has a number of distinct characteristics which determine its

usability and potential value and applications. First, based on the review above,

blockchain is a completely decentralized system. Because it relies on a distributed

network to conduct and confirm transactions, no third parties are involved, and no single

centralized structure is necessary to its use. All transactions can be conducted from

different systems and devices connected to a network (Chen et al., 2018).

Second, blockchain ensures the absolute transparency of all transactions. They are

chronologically arranged with unique timestamps. Any node on the network has access

to the same general ledger. It can inspect any block and transactions that the block

describes. This not only ensures openness and equal access to transaction-related

information but also eliminates the need for such a property as trust. Indeed, there is no

need to trust another party in a transaction when the entire history can be freely

examined and confirmed with no possibility for anyone to singlehandedly alter the

existing records (Abeyratne & Monfared, 2016).

Third, blockchain ensures the immutability of transactions. As previously discussed, the

features which are unique to each transaction described by encrypted hash links,

timestamps, and nonces make it impossible to tamper with transactions. One way to

tamper with blockchain would be to achieve a simultaneous change to the distributed

ledger across the majority nodes on the network, which is essentially impossible due to

complexity and virtually unattainable required technological power (Chen et al., 2018).

Finally, blockchain can be thought of in terms of cryptocurrency. Every network

utilizing blockchain involves direct node-to-node transactions with a fixed circulation

property which is defined by a strict mathematical algorithm (Chen et al., 2018). In such

a system, no transactions are lost, and no system collapse of inflation will take place.

This makes the system stable, pre-defined, and capped based on the originally given

properties. For example, the original blockchain algorithms capped the amount of

Bitcoins at 21 million (Nakamoto, 2008).

While the original use of blockchain was envisioned as a cryptocurrency driver, the

properties described above make it a potentially desired technology for other

applications. These were presented in a major work by Buterin (2014) which introduced

a novel concept of Etherium blockchain with an embedded open source programming

language. Etherium granted the possibility for everyone to write smart contracts and

also opened doors for various applications of blockchain. Today, researchers indicate the

current state of blockchain technology development as Blockchain 3.0 (Gatteschi,

Lamberti, Demartini, Pranteda, & Santamaria, 2018). It is characterised by novel

applications of blockchain in non-financial sectors. Higher education is one such field.

1.5 Research Significance and Contribution

This research operates on the intersection of the technology adoption and education

improvement studies in the Saudi Arabia context. Considering the increased student

mobility and globalised nature of higher education, colleges and universities in Saudi

Arabia are faced with growing competitive pressures. Innovative technologies in

education have long ago been recognised as an important factor in gaining competitive

advantage (Cheng, Cham, Dent, & Lee, 2019; Mainardes, Ferreira, & Tontini, 2011;

Waller, Lemoine, Mense, & Richardson, 2019). This motivates higher education

institutions to constantly look for new opportunities by adopting novel technologies

such as blockchain.

Despite the relative newness of blockchain, it has already found practical applications in

a number of industries, such as banking and finance (Kulkarni & Patil, 2020; Rajnak &

Puschmann, 2020), supply chain management (Alazab, Alhyari, Awajan, & Abdallah,

2021; Aslam, Saleem, Khan, & Kim, 2021; Choi, Chung, Seyha, & Young, 2020),

energy (Andoni, et al., 2019; Wang & Su, 2020), the public sector (Reddick, Cid, &

Ganapati, 2019; Warkentin & Orgeron, 2020), healthcare (Hasselgren, Kralevska,

Glikoroski, Pedersen, & Faxvvag, 2020; Pirtle & Ehrenfeld, 2018), and tourism

(Rashideh, 2020; Valeri & Baggio, 2021) among others. The potential of blockchain in

education is closely related to the continuing digitization of educational services, which

requires an increased level of security, speed and reliability of information exchange,

enhanced data management, and prevention of fraud. Many of these and other pertinent

issues in education and related services can be potentially addressed by the unique

features of blockchain technology such as decentralization, transparency, immutability,

and security of transactions (Alammary, Alhazmi, Almasri, & Gillani, 2019; Raimundo

& Rosario, 2021). This explains the great amount of interest in blockchain for education

from both academics and technology practitioners. However, researchers have noted

that literature in this field remain fragmented, focusing primarily on existing and

potential blockchain applications as well as opportunities and challenges for these

applications in education institutions (Alammary, Alhazmi, Almasri, & Gillani, 2019;

Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020).

The limited and incomplete body of research on blockchain adoption in education in

general and in Saudi Arabia universities in particular serves as the primary motivator for

this study. The expected contribution of this research is both theoretical and practical as

outlined below.

1.5.1 Theoretical Contributions

a. This study extends the technology adoption theory by identifying factors

influencing blockchain computing adoption in the Saudi education sector. It

expands on the well-known theories of adoption by integrating several

wellknown theoretical constructs into a more holistic framework.

b. The study presents a refined theory of blockchain adoption in the higher

education sector and tests it within a specific context where it has not been tested

before. As such, the role of both theoretically established and newly discovered

contextual factors is explored.

c. The study tests the applicability of the well-known adoption and established

theories in the Saudi context.

1.5.2 Practical Contributions

a. The study produces an actionable framework of blockchain adoption which

demonstrates what factors influence and impede blockchain adoption in higher

education institutions in Saudi Arabia.

b. The study offers a roadmap for Saudi colleges’ and universities’ decision makers

in implementing blockchain-related initiatives within their respective

institutions.

c. The study offers a validated instrument (questionnaire) to study adoption either

through a combination of the considered factors or test the effect of each

dimension more thoroughly.

d. To the best knowledge of the author, this is the first study of its kind to

empirically explore the combinatory effect of technological, organisational,

environmental and quality factors and barriers to blockchain adoption in Saudi

HEIs. It can serve as a foundation for developing a tool to guide HEI

organisations and, perhaps, the industry as a whole on whether, when and how

the adoption process should proceed.

1.6 Thesis Outline

This thesis consists of nine chapters, described as follows:

Chapter 1 provides a general research background, including the study context and the

blockchain technology basics. It also explains the current research’s significance and

contribution.

Chapter 2 provides a systematic literature review of blockchain adoption in education.

Three major areas are covered: evidence of blockchain adoption in higher education,

reasons for blockchain adoption in higher education and relevant factors in the adoption

process. A rigorous methodology for the search and inclusion of studies in the review is

discussed, and the analysis of themes for blockchain and potential uses in higher

education is presented. The findings are used to identify the gaps in the existing

knowledge to place this research within the body of knowledge on blockchain adoption

in higher education.

Chapter 3 presents the problem definition, outlines the research questions and

objectives.

Chapter 4 provides an overview of the solution to the formulated problem, which is split

into research subquestions. This chapter also reviews the study methodology and

describes the major research phases.

Chapter 5 presents the research model to study blockchain adoption in Saudi HEIs. Each

model dimension is discussed separately and the relationships between the major factors

are presented and justified. Accordingly, the key hypotheses to investigate within the

framework are presented.

Chapter 6 offers a detailed review of the Phase I results of the research. Phase I is

represented by a qualitative data collection and analysis. Accordingly, the chapter

provides an overview of the research sample and the results of the interviews. Based on

these results, the original model presented in Chapter 4 is refined to become the

foundation for the wider quantitative research.

Chapter 7 offers a detailed review of the Phase II results of the research. These are the

results of a large-scale survey. The chapter first offers a descriptive analysis of the study

sample. Next, preliminary analyses of the data are reported, including tests for data

normality, validity, reliability, and multicollinearity. Adjustments to the model are made

and justified where necessary. Finally, the results of the hypotheses tests are presented.

Chapter 8 offers a comprehensive review and discussion of the study findings. It begins

with a review of the study model evolution and a comparison of the results from Phase I

and Phase II of the research. Next, thorough discussions and explanations of the results

are provided for every hypothesised relationship in the study model.

Chapter 9 draws the main conclusions arising from the study. It places the study findings

within the existing body of research on blockchain adoption in education and draws

theoretical and practical implications of the research. Finally, general limitations of the

study are acknowledged and directions for future research are outlined.

Figure 5 visualises the study plan and shows how the chapters are connected to each

other.

1.7 Conclusion

This chapter provided an overview of the work undertaken within the current thesis.

Given the importance of technological developments in the education sector, the study

looks into the process of blockchain adoption in the context of Saudi colleges and

universities. The purpose is to develop and test a practical framework which could be

used by the administrators to guide the adoption process in an effective way. The study

contribution, therefore, is both theoretical and practical. On the one hand, it proposes a

new framework on the basis of the existing adoption theories, supplements it with the

contextual factors and empirically tests the proposed relationships. On the other hand,

the resulting framework is an actionable tool which can be applied in practice to

stimulate blockchain adoption in higher education establishments.

Setting research context, outlining

contributions

Summarizing existing knowledge,

identifying gaps

Research question, problems addressed

Blockchain Adoption Model development

1.Introduction

2.Literature Review

.Solution Overview4

3.Problem Definition

.Blockchain5

Adoption Model and

Hypotheses

Defining model relationships and approaches to test them

Interviews, model refinement

Survey finalized model testing.

Review of the study findings

Implications of the findings, future research propositions

Figure 5: Thesis Structure

2 A Systematic Literature Review of Blockchain Adoption in

Education

2.1 Introduction

The purpose of this chapter is to review the relevant literature pertaining to the research

topic. Three major areas are covered: evidence of blockchain adoption in higher

education, reasons for blockchain adoption in higher education and relevant factors in

the adoption process. A rigorous methodology for the search and inclusion of the studies

in the review is discussed, and the analysis of themes for blockchain and its potential

uses in higher education is presented. The findings are used to identify the gaps in the

existing knowledge to place this research within the body of knowledge on blockchain

adoption in higher education.

Section 2.2 defines the method used for extracting and analysing the literature. Section

2.3 summarises the extracted literature by year and publication type. Section 2.4 reviews

the existing evidence of blockchain adoption in the educational sector. Section 2.5

summarises the key reasons for higher education institutions (HEIs) to adopt

blockchain. Sections 2.6 and 2.7 review the adoption barriers and factors respectively as

identified in the existing literature. Section 2.8 reviews the theoretical frameworks for

blockchain adoption in HEIs. Section 2.9 discusses the key limitations of the conducted

literature review. Finally, Section 2.10 outlines the major research gaps arising from the

literature review.

:.Phase I Research6

Qualitative

7.Phase II Research :

Quantitative

8.Discussion of

Results

9.Conclusion and

Future Work

Sections of this chapter have earlier been published in the following journal article:

Alalyan, M.S., Jaafari, N.A., Hussain, F.K. & Gill, A.O. (2023). A systematic review of

blockchain adoption in education institutions. International Journal of Web and

Grid Services,19(2), 156-184.

2.2 Literature Research Method

This study followed a systematic literature review (SLR) approach. An SLR is generally

defined as “a literature review that is designed to locate, appraise, and synthesise the

best available evidence relating to a specific research question in order to provide

informative and evidence-based answers” (Boland, Cherry, & Dickson, 2017, p. 2). It is

recognised as more encompassing and rigorous in comparison to the traditional

approach to a literature review based on narratives (Booth, Papaioannou, & Sutton,

2012; Okoli, 2015).

by education institutions? blockchain adoption by education institutions.

RQ4: what are the barriers to blockchain adoption To determine and classify the factors that impede by

education institutions? blockchain adoption by education institutions.

RQ5: what theoretical frameworks have been used To identify the models that can be useful for future

to study blockchain adoption by education empiric investigations of blockchain adoption. institutions?

The literature search was conducted across eight scientific databases:

•ACM

•IEEE

•ProQuest

•EBSCO IT

•ScienceDirect

•SpringerLink

•Taylor & Francis

•Web of Science

Additionally, two education research databases were searched:

•ERIC

•Education Research Complete.

The databases were selected for their reputation, high index impact, and specialization

in technology and education topics. Additionally, a search by Google Scholar was

conducted at the end to account for publications potentially missed during the original

search. The time frame was set between 2008 (first paper published on blockchain by

Nakamoto) and May 2021. Following Okoli (2015) and Kitchenham & Charters (2007),

the search strings were developed based on the research questions’ themes, alternative

spellings of the key terms, and using the Boolean operators. The following query strings

were used:

[“Blockchain” OR “Block chain” OR “Distributed Ledger”] AND [“Application” OR

“Use” OR “Usage”] AND [“Education” OR “Learning” OR “Teaching”]

[“Blockchain” OR “Block chain” OR “Distributed Ledger”] AND [“Adoption” OR

“Use” OR “Usage”] AND [“Education” OR “Learning” OR “Teaching”]

2.2.2 Step 2: selection

After the search, duplicate titles and papers were eliminated from the review with the

most recent version retained. Next, the titles and abstracts of the retained papers were

screened for the initial inclusion of articles in the review. The following exclusion

criteria were applied:

1. the paper is not in English;

2. the paper is incomplete or missing full text (in press articles were still

included); 3. the paper does not cover blockchain adoption and/or use by

education institutions;

4. the paper represents an opinion without solid methodology or research.

2.2.3 Step 3: quality assurance

The full textы of the remaining articles were screened to ensure that they were relevant

to the research questions and represented high quality research (quality assurance). The

papers were retained for analysis based on a quality score derived from four quality

assurance questions listed in Table 2. The answers received the following grading: 1

point for “Yes,” 0.5 point for “Partially,” and 0 for “No.” The acceptable score for

inclusion in the analysis was 2.5 and higher.

Table 2: Quality Assurance Questions

Quality Assurance Questions Grading Criteria

1. Do the authors state

the study purpose?

2. Is the methodology

clearly described?

3. Is the level of rigor

for study execution and

analysisappropriate?

4. Are the findings

useful for academic

research and/or

practicaluse?

Yes = 1

Partially = 0.5

No = 0

2.2.4 Step 4: execution

The final step in the review involved data synthesis and reporting the results. The data

about and from the extracted papers were summarised and grouped into themes

corresponding to the research questions. The final results are presented in the form of

tables and figures for ease of analysis and interpretation.

2.3 Summary of the Extracted Literature

The initial searches over the databases returned 335 studies in total: 124 from Scopus, 65

Scopus: 124

IEEE: 65

ProQuest: 62

EBSCO

Education: 59

ACM: 15

EBSCO IT: 10

Publication Type Number of Papers

Conference Paper

Journal Article

Book Chapter

Workshop Paper

Other (report, colloquium, symposium) 3

Total 107

was performed. This enabled the development of a taxonomy of the reviewed

literature. In research, a taxonomy is generally understood as a structured and organised

presentation of information and knowledge (Yazdani, Shirvani, & Heidarpoor, 2021). A

number of approaches to taxonomy development for reviewed literature exist (e.g.,

(Creswell, 2015; Della Porta & Keating, 2008; Gall, Gall, & WR, 2007)). While

different in terms of the classification of studies, they share similar methods of

organising knowledge based on either research methodology or purpose of research. This

study followed this pattern. The taxonomy of the reviewed research is presented in

Figure 8. Three general research types were identified: 1) reviews of the existing

literature and cases of blockchain use in education; 2) conceptual models and proposals

for blockchain solutions and applications; and 3) empirical studies of factors influencing

blockchain adoption in education. Within each of these types, research subcategories

were distinguished based on the study foci, research questions, and purpose. Detailed

analyses for each research type dimension are presented in Figure 8.

Initial Search: 335 studies retrieved

Screening (duplicate removal): 241 studies

remained

Screening (abstract analysis): 161 studies

remained

Quality Assurance (in-depth analysis): 107 studies

remained

Figure 8: Blockchain Adoption in Education: Research Taxonomy

2.4.1 Literature Surveys

In total, 32 papers were identified as literature surveys, which represents 29.9% of all

studies. Several types of reviews were identified, which is indicated in Table 4. Each is

discussed in Table 4.

Studies of Blockchain Adoption in Education

Literature Surveys

Existing Case Uses

Potential Applications

General Reviews

Focused Reviews

Models

Empirical Research

Conceptual

Tested, not applied

Working

Quantitative Research

Qualitative Research

Table 4: Subcategories of Reviews of Blockchain Adoption in Education

Subcategory Number Studies

Reviews of Use Cases 7 (Bhaskar, Tiwari, & Joshi, 2020), (Jirgensons &

Kapenieks, 2018), (Sharma & Batth, 2020), (Capece,

Ghiron, & Pasquale, 2020), (Fedorova & Skobleva,

2020), (Guustaaf, Rahardja, Aini, Maharani, &

Santoso, 2021), (Hameed, et al., 2019), (Kamisalic,

Turkanovic, Mrdovic, & Hericko, 2020)

Reviews of

Applications

Potential 5 (AlHarthy, AlShuhaimi, & AlIsmaili, 2019), (Chen, Xu,

Lu, & Chen, 2018), (Lindenmoyer & Fischer, 2019),

(Haugsbakken & Langseth, 2019), (Kant & Anjali,

2020)

General Surveys 9 (Bhaskar, Tiwari, & Joshi, 2020), (Alammary,

Alhazmi, Almasri, & Gillani, 2019), (Raimundo &

Rosario, 2021), (Awaji, Solaiman, & Albshri, 2020),

(Loukil, Abed, & Boukadi, 2021), (Machado, Sousa, &

Rocha,

2020), (Yue, Xiaofeng, & Huagang, 2020), (Gresch &

Camilleri, 2017), (Yumna, Khan, Ikram, & Ilyas, 2019)

Focused Surveys 11 (Ma & Fang, 2020), (Novotny, et al., 2018), (Williams,

2019), (Arndt & Guercio, 2020), (Caldarelli & Ellul,

2021), (Castro & Au-Yong-Oliveira, 2021),

(Fernandes-Carames & Fraga-Lamas, 2019), (Liu &

Zhu, 2021), (Mikroyannidis, Domingue, Bachler, &

Quick, 2018), (Pfeiffer, Bezzina, Wernbacher, &

Kriglstein, 2020), (Sahonero-Alvarez, 2018)

Reviews of use cases are papers which discuss real-life applications of blockchain in

education. In total, 7 such papers were identified. Four studies described use cases by

educational institutions: colleges and universities. Capece et al. (2020) reviewed uses

cases of Blockcerts developed by Massachusetts Institute of Technology (MIT).

Jirgensons and Kapieniks (2018) reviewed blockchain applications by five universities

in USA and Europe. More comprehensive reviews by Kamisalic et al. (2020) and

Fedorova and Skobleva (2020) respectively identified use cases in 20 and 23 colleges

and universities globally. The total number of blockchain adopting education institutions

described in these studies is 29. The majority of the institutions adopted blockchain for

issuing, storing, and verifying diplomas. Six institutions used blockchain for student

identity verification and preventing fraud. Five institutions integrated payments in

cryptocurrencies. Five colleges and universities used blockchain solutions to create

novel forms of online education interactions between students and instructors. Three

institutions, all Chinese, used blockchain for intellectual property protection. Two

universities integrated blockchain for administrative tasks such as the management of

digital microcredentials. One university experimented with a blockchain-based

accreditation system, and one university introduced a blockchain-based peer review

academic publishing platform. Table 5 offers a summary of the use cases discussed in

literature.

Table 5: Blockchain Adoption by Colleges and Universities

Application Field N Education Institutions

Issuing, storing, and verifying

diplomas

22 Aristotle Athens University of Economics and Business, Central

New Mexico Community College, Democritus University of

Thrace, Holberton School of Software Engineering, Malta College

of Arts Science and Technology, MIT, Ngee Ann Polytechnic,

Penza State University, Southern New Hampshire University,

Synergy University, The Open University, University of Bahrain,

University of California, University of Maribor, University of

Melbourne, University of New Hampshire, University of Nicosia,

University of Rome, University of Southampton, University of

Texas at Austin,

University of Thessaloniki, Woolf University

Student identity management and

solutions

6 Aristotle Athens University of Economics and Business,

Democritus University of Thrace, Holberton School of Software

Engineering, MIT, University of Thessaloniki, Woolf University

Cryptopayments 5 King’s College, Simon Fraser University, University of Cumbria,

University of Nicosia, Woolf University

Novel teacher/learner platforms 5 Synergy University, University of Southampton, The Open

University, University of Texas at Austin, Woolf University

Intellectual property protection 3 Chinese Academy of Sciences, Shenzhen University, Zhejiang

University

Administrative tasks 2 University of Maribor, Woolf University

Institution accreditation 1 The Open University

Peer review open publishing

platform

1 University of Pittsburgh

It should be noted, however, that the actual number of educational institutions that either

adopted or experimented with the adoption of blockchain is likely much higher. For

example, Fedorova and Skobleva (2020) reported that up to 20% of Canadian colleges

and universities could be using blockchain technology for education achievement

certificates and diplomas. Therefore, there is a growing need to continuously and

consistently report on blockchain adoption in higher education to have a clearer picture

of its current state and perspectives in the industry.

Another three papers focused on blockchain-based education platforms. Sharma and

Bhutt (2020) described five such platforms. Hameed et al. (2019) discussed nine.

Finally, Guustaaf et al. (2021) reviewed twelve ready-to-use blockchain platforms for

the educational sector. However, these papers were mostly descriptive. While offering

useful and comparative reviews of the blockchain-for-education projects, they did not

discuss their uses by educational institutions. As such, there is no academic evidence of

the adoption of such systems or the success of these initiatives.

A separate category of reviews considered potential applications of blockchain for

education. These were the papers where researchers reviewed the areas in education

where blockchain could offer benefits for both institutions and students. In total, 6 such

papers were identified, which are summarised in Table 6. Overall, the range of potential

applications for blockchain is rather wide, which presents many areas for future research

and case studies on novel approaches to blockchain use in education.

Table 6: Papers on Potential Blockchain Uses in Education

Paper Potential Application Fields Reviewed

Chen et al. (2018) Achievements and certificates, assessment platforms, online learning, digital

badges, smart contracts

Al Harthy et al. (2019) Digital badges, direct transactions, secured records, library administration,

instructor achievements, publishing

Lindenmoyer & Fischer

(2019)

Timeless achievement data, prevention of academic fraud, secure and trusted

resumes, admissions, academic advancement tracking

Haugsbakken et al. (2019) Transcripts management, competence badges, learner digital identity, flexible

degree design, secure intellectual property

Kant & Anjali (2020) Administrative tasks, learning delivery, record keeping, accreditation,

transfers, digital badges, smart contracts

Nurhaeni et al. (2018) Academic certificate management, open badges

The third type of literature survey papers are general reviews. These cover a broad range

of topics on blockchain without focusing on particular blockchain applications or areas.

In total, 9 papers were identified. The unifying purpose of such reviews is usually to

offer a comprehensive view on the state of research. It involved an examination of the

blockchain applications in education, developing a list of challenges and benefits of

blockchain for education and identifying the research gaps. At the same time, some

topics were omitted in these reviews. No paper attempted to systematise the existing

research by type and develop a taxonomy of knowledge. Further, only one paper

reviewed the potential factors driving the adoption of blockchain by education

institutions. These are the clear gaps that this research aims to fill. Table 7 summarises

general review studies and the topics they covered on a comparative basis with this

research.

Table 7: General Reviews of Blockchain Adoption in Education

Study SLR? Applications

Reviewed

Reasons to

Adopt

Reviewed

Barriers to

Adopt

Reviewed

Factors Driving

Adoption

Reviewed

Taxonomy

Developed

Gresch & Camilieri (2017) x x

Alammary et al. (2019) x x X x

Yumna et al. (2019) x x X x

Awaji et al. (2020) x x x

Bhaskar et al. (2020) x x X x

Machado et al. (2020) x x X x

Yue et al. (2020) x X x x

Loukil et al. (2021) x x X x

Raimundo & Rosario

(2021)

x x X x

This Research x x X x x x

Focused reviews represent the largest proportion of literature surveys with 11 papers in

total. These papers explored the state of blockchain adoption in a particular field or area

of education. The number of focused surveys increased rapidly since 2018 which

coincides with the reports of successful implementations of blockchain in a number of

universities and possibly with the need to look into new areas of blockchain

applications. The majority of the reviews focused on blockchain applications for student

records and proof of education. Arndt and Guercio (2020) and Caldarelli and Ellul

(2021) reviewed blockchain applications for transcript management. Castro and Au-

Yong-Oliveira (2021) reviewed blockchain applications for issuing and verifying

diplomas. A review by Ma and Fang (2020) focused on blockchain applications for

record keeping, decentralised education, and certificates. Pfeiffer et al. (2020) reviewed

blockchain applications for student data management and the prevention of identity

fraud. Three reviews focused on blockchain applications for the creation of new,

student-centric online learning environments. Mikroyannidis et al. (2018) reviewed

literature related to blockchain for eportfolios, accreditation, and tutoring. Williams

(2019) considered blockchain applications for decentralised, individualised learning

curricula. Fernández-Caramés (2019) reviewed blockchain applications for the creation

of a smart university. Two studies offered focused reviews on blockchain applications

for specific learning programs such as engineering education (Sahonero-Alvarez, 2018)

and cultural, creative design (Liu & Zhu, 2021). Finally, Novotny et al. (2018) reviewed

blockchain literature related to blockchain applications in academic publishing.

One clear limitation of these focused reviews, however, is the absence of rigorous

literature review methodologies. In fact, only Caldarelli & Ellul (2021) and Castro &

AuYong-Oliveira (2021) used systematic literature review approaches that would enable

their research to be replicated and expanded. It is also clear that the number of potential

education fields for blockchain applications represented in these studies is rather limited

in comparison to the range of applications described in the general review studies and

discussed above.

2.4.2 Models

The second largest category of studies of blockchain adoption in education is

represented by model studies with a total of 69 papers or 64.5% of the total number.

Three types of papers in this category can be identified: conceptual models, tested

models, and working models.

Conceptual models are the largest subcategory with 41 papers reviewed. These studies

are proposals for blockchain integration into various aspects of education process. The

authors would normally present a model/architecture, discuss its benefits and the

solutions they could provide in specific areas. The majority of conceptual proposals

were in the areas of certification and degrees and administrative processes. Fewer, but

still a substantial number of researchers proposed models for enhancing the learning

process and outcome assessments. Models for intellectual property protection and smart

universities were described by one paper each. The blockchain application categories

discussed in the conceptual papers are presented in Table 8.

Table 8: Conceptual Models for Blockchain Implementation in Education

Application

Category

School

Certificates

Certificate issue and

verification, Diploma issue

and verification,

(Budhiraja & Rani, 2019),

(Gräther, et al., 2018), (Liu,

Xiao, Tang, & Hosam, 2020),

(Palma, Vigil, Pereira, &

Martina, 2019),

6 30.0%

(Vidal, Gouveia, & Soares,

2019), (Xu, et al., 2017)

Administration Transcript management,

Admissions, Education

records, Grade storage, Data

security, Cryptopayments,

Authentication, Smart

contracts

(Arndt & Guercio, 2020),

(Hori & Ohashi, Adaptive

Identity authentication of

blockchain system-the

collaborative cloud educational

system, 2018), (Ismail,

Hameed, AlShamsi,

AlHammady, & Aldhandhani,

2019), (Kanan, Obaidat, &

AlLahham, 2018), (Mori &

Miwa, 2019), (Rooksby &

Dimitrov, 2019)

6 30.0

Enhanced

Learning

Environments

Systems for learning

evaluation and rewards,

Curriculum design and

personalization, Student

accreditation,

(Bdiwi, de Runz, Cherif, & Faiz,

2019), (Kontzinos, et al., 2019),

(Lam & Dongol, 2020),

(Sharples & Domingue, 2016)

4 15.0

Learning and

Assessments

Lifelong learning, learning

trace repositories, Exam

paper distribution and audit

(Cahyadi, Faturahman,

Haryani, Dolan, & Millah,

2021), (Farah, Vozniuk,

Rodriguez-Triana, & Gillet,

2018), (Mitchell, Hara, &

Sheriff, 2019), (Ocheja,

Flanagan, Ueda, & Ogata, 2019)

4 15.0

20 100.0%

The final subcategory includes working models: those which have been successfully

implemented at educational institutions on a constant basis. In total, 7 case study papers

were identified. The earliest study was published by Bore et al. (2017) and described the

design, implementation, and evaluation of a blockchain-enabled school information hub

in Kenya. Hori et al. (2018) described the CHiLO project – a decentralised learning

system using virtual currency. The authors discussed the successful implementation of

the first project phase for the creation and publication of e-books. Curmi and Iguanez

(2018) presented action research on a prototype blockchain-based model for academic

certificate issue and verification. Turkanovic et al. (2018) described EduCTX – a

blockchain-based credit platform which has been implemented in two Slovenian

universities. Vidal et al. (2019) described a system for blockchain-based diplomas at

University Fernando Pessoa, Portugal, although they admitted that the system was

feasible only if a larger number of blockchain-based diplomas are issued due to

economies of scale. Mahankali and Chaudhary (2020) described the successful

application of AuxCert – a blockchain-based certificate management platform for a

university in India. Finally, Dudhat et al. (2021) described the Edublocs project at the

University of Barcelona used to oversee and record students’ academic activities.

2.4.3 Empirical Research on Adoption Factors/Barriers

The final category of studies on blockchain for education consists of empirical

investigations into the factors that enable and/or impede blockchain adoption. This is by

far the smallest category of studies with only 5 papers identified (Table 10). Of these, 4

papers had qualitative research designs based on interviews and focus groups. Only one

paper (Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020) implemented a

quantitative study based on a survey. Three of these papers explored the factors that

could stimulate blockchain adoption by education institutions, while two focused

exclusively on barriers to adoption.

Table 10: Empirical Studies of Blockchain Adoption in Education

Study Research Design Factors Influencing

Adoption

Barriers to

Adoption

Fedorova & Skobleva (2020) Qualitative: interviews No Yes

Kosmarski (2020) Qualitative:

interviews and

focus groups

No Yes

El Nokiti and Yusof (2019) Qualitative: focus groups Yes No

Ullah et al. (2020) Quantitative: survey Yes No

Widjaja et al. (2020) Qualitative: interviews Yes No

Several important observations arise from the review. On a positive side, there seems to

be a growing interest in blockchain applications for education. The studies covered a

wide range of applications, both existing and potential. There is also growing evidence

of working blockchain models for education institutions with dozens adopting and

applying blockchain solutions for various purposes. At the same time, the research

remains dominated by review papers and conceptual models. There is a limited number

of papers outside the major research areas, which are certificates/diplomas and

blockchain for administrative purposes such as admissions, management of transcripts

and student records. Blockchain applications for learning processes seem to be gaining

interest, although the number of such papers remains disproportionately small. As such,

the research remains misbalanced in terms of topics and conceptual versus real-life

applications. Finally, it is clear that the field lacks empirical research and case analyses

of blockchain adoption in education. In the absence of strong empirical evidence of

blockchain benefits for different education areas as well as the processes behind

adoption and the factors driving it, many institutions may remain cautious regarding

blockchain.

2.5 Reasons to Adopt Blockchain in Education

Blockchain, as any other novel technology, has been thoroughly analysed by the

researchers to identify the reasons for its adoption in education. In essence, studies

offering an analysis of reasons to adopt answer the question as to why blockchain

should be integrated into education environments. In total, 33 papers in the review listed

specific adoption reasons. The discussions of blockchain adoption reasons commonly

follow reviews of typical problems faced by educational institutions today. Among these

are: problems associated with physical credential confirmations and single points of

failure; problems associated with academic data storage and exchange; problems

associated with academic fraud and others (Alammary et al., 2019; Chen et al., 2018;

Kamisalic et al., 2020; Gresch & Camillieri, 2017; Yumna et al., 2019). Accordingly,

researchers view the benefits of blockchain adoption for both education institutions and

students. A summary of these is presented in Table 11.

Table 11: Reasons to Adopt Blockchain in Education Identified in Literature

Reasons Adopt to Specific Benefits Studies

Organisationfocused -Reduced cost of operations

-Reduced cost of

data management

-Reduced administrative

personnel requirements

-Automation of processes

-Enhanced data security

-Reduced bureaucracy

(Awaji, Solaiman, & Albshri, 2020), (Ma &

Fang, 2020), (Holotescu, 2018)

(Bhaskar, Tiwari, & Joshi, 2020), (Kant &

Anjali, 2020), (Eaganathan, Indrian, &

Nathan, 2019), (Liu, et al., 2021), (El

Nokiti & Yusof, 2019)

(Alammary, Alhazmi, Almasri, & Gillani,

2019), (Lindenmoyer & Fischer, 2019),

(Yumna, Khan, Ikram, & Ilyas, 2019),

(Turkanovic, Holbl, Kosic, Hericko, &

Kamisalic, 2018)

(Haugsbakken & Langseth, 2019)

(Bhaskar, Tiwari, & Joshi, 2020),

(Alammary, Alhazmi, Almasri, & Gillani,

2019), (AlHarthy, AlShuhaimi, &

AlIsmaili, 2019), (Jirgensons & Kapenieks,

2018), (Abreu, Coutinho, & Bezerra,

2020), (Eaganathan, Indrian, & Nathan,

2019), (Liu, et al., 2021), (El Nokiti &

Yusof, 2019)

(Haugsbakken & Langseth, 2019)

Student-focused -New methods of course delivery

and assessment

-Stronger collaborative

studentstudent and student-

instructor environments

-Better organization of

(Alammary, Alhazmi, Almasri, & Gillani,

2019), (Williams, 2019), (El Nokiti &

Yusof, 2019)

(Loukil, Abed, & Boukadi, 2021),

(Novotny, et al., 2018)

knowledge and learning process

-“Learning is earning” approach

-Reduced cost of education for

students

(Raimundo & Rosario, 2021)

(Chen, Xu, Lu, & Chen, 2018),

(SahoneroAlvarez, 2018), (Lizcano, Lara,

White, & Aljawarneh, 2020), (Kontzinos,

et al., 2019)

(Kant & Anjali, 2020), (Castro & Au-

YongOliveira, 2021)

Organization and

student focused

-Immutable, easily

verifiable academic

credentials

-Reduced academic fraud

-Enhanced privacy and data

ownership

(Bhaskar, Tiwari, & Joshi, 2020),

(Raimundo & Rosario, 2021), (Loukil,

Abed, & Boukadi, 2021), (Yumna, Khan,

Ikram, & Ilyas, 2019), (Arndt & Guercio,

2020), (Abreu, Coutinho, &

Bezerra, 2020), (Lizcano, Lara, White, &

Aljawarneh, 2020), (Abougalala, Amasha,

Areed, Alkhalaf, & Khairy, 2020),

(Ocheja, Flanagan, Ueda, & Ogata, 2019)

(Alammary, Alhazmi, Almasri, & Gillani,

2019), (Castro & Au-Yong-Oliveira,

2021), (Chen, Xu, Lu, & Chen, 2018),

(Widjaja, Cassandra, Widjaja, Prabowo, &

Fernando, 2020), (Nurhaeni, Handayani,

Budiarty, Apriani, & Sunarya, 2020)

(Bhaskar, Tiwari, & Joshi, 2020),

(Alshahrani, Beloff, & White, 2020), (Lee

& Park, 2021), (Abougalala, Amasha,

Areed, Alkhalaf, &

Khairy, 2020), (El Nokiti & Yusof, 2019)

From a purely organisational standpoint, blockchain is often associated with operational

efficiencies and improvements of administration processes. Reducing the costs of

operations is often related to the distributive and immutable nature of blockchain

(Awaji, Solaiman, & Albshri, 2020; Holotescu, 2018; Ma & Fang, 2020). It enables

costs to be saved on data management (Bhaskar et al., 2020; Eaganathan et al., 2019; El

Nokiti & Yusof, 2019; Kant & Anjali, 2020; Liu et al., 2020), a reduction in the required

administrative personnel (Alammary et al. 2019; Lindenmoyer & Fischer, 2019;

Turkanovic et al., 2018) and the automation of certain operations (Haugsbakken &

Langseth, 2019). Finally, blockchain is often regarded as a next step in data security,

which is more efficient and reliable than the existing database approaches (Abreu et al.,

2020; Alammary et al., 2019; AlHarthy et al., 2019; Bhaskar et al., 2020; Eaganathan et

al., 2019; El Nokiti & Yusof, 2019; Jirgensons & Kapenieks, 2018; Liu, et al., 2021).

In terms of student-centred advantages, researchers pointed out to blockchain’s ability to

improve the overall quality of education and contribute to contemporary education

development (Cahyadi et al., 2021; Nurhaeni et al., 2020). Blockchain is regarded as a

means to enhance learning environments through new methods of course delivery and

assessment (Alammary et al., 2019; El Nokiti & Yusof, 2019; Williams, 2019),

facilitating stronger collaborative student-student and student-instructor environments

(Loukil et al., 2021; Novotny, et al., 2018) and better organisation of knowledge and

learning process (Raimundo & Rosario, 2021). An important related innovation offered

by blockchain is the concept of “learning is earning” where institutions can create a

form of cryptocurrency to track student progress and offer rewards (Chen, Xu, Lu, &

Chen, 2018; Kontzinos, et al., 2019; Lizcano, Lara, White, & Aljawarneh, 2020;

SahoneroAlvarez, 2018). Several authors also noted an opportunity to reduce the cost of

education for students (Castro & Au-Yong-Oliveira, 2021; Kant & Anjali, 2020).

Finally, a wide array of benefits from adopting blockchain in education falls into the

mutual category for students and organizations. They are often linked to a new, unique

type of trust mechanism that blockchain can offer (Caldarelli & Ellul, 2021; Novotny et

al., 2018). For example, the distributed nature of blockchain is also considered a great

contribution to the security and transparency of academic achievements and credentials.

Specifically, researchers pointed to immutable blockchain-secured diplomas which can

be easily verified and shared between education institutions and employers (Abougalala

et al., 2020; Abreu et al., 2020; Arndt & Guercio, 2020; Bhaskar et al., 2020; Lizcano et

al., 2020; Loukil et al., 2021; Ocheja et al., 2019). Reducing academic fraud and

stronger identity authentication are also considered important reasons to adopt

blockchain in education (Alammary, Alhazmi, Almasri, & Gillani, 2019), (Castro & Au-

Yong-Oliveira, 2021), (Chen, Xu, Lu, & Chen, 2018), (Widjaja, Cassandra, Widjaja,

Prabowo, & Fernando, 2020), (Nurhaeni, Handayani, Budiarty, Apriani, & Sunarya,

2020). Likewise, researchers noted that a beneficial feature of blockchain for both

students and institutions is the ability to enhance privacy and data ownership (Bhaskar,

Tiwari, & Joshi, 2020), (Alshahrani, Beloff, & White, 2020), (Lee & Park, 2021),

(Abougalala, Amasha, Areed, Alkhalaf, & Khairy, 2020), (El Nokiti & Yusof, 2019).

2.6 Adoption Barriers

Barriers to blockchain adoption in education have been well explored in the literature.

These barriers can be categorised as: technology related, organization related, and

environment related.

Technology-related barriers to blockchain adoption have been discussed in terms of

innate blockchain features and technology novelty (Table 12). Scalability of blockchain

was the most often mentioned issue, as the existing blockchains demonstrate reduced

performance levels for continuously expanding data (Bhaskar, Tiwari, & Joshi, 2020),

(Alammary, Alhazmi, Almasri, & Gillani, 2019), (Awaji, Solaiman, & Albshri, 2020),

(Loukil, Abed, & Boukadi, 2021), (Ma & Fang, 2020), (Sharma & Batth, 2020), (Yue,

Xiaofeng, & Huagang, 2020), (Yumna, Khan, Ikram, & Ilyas, 2019), (Caldarelli & Ellul,

2021), (Abougalala, Amasha, Areed, Alkhalaf, & Khairy, 2020), (Ismail, Hameed,

AlShamsi, AlHammady, & Aldhandhani, 2019), (Ocheja, Flanagan, Ueda, & Ogata,

2019). Transaction speed and processing time are seen as problems related to this (Chen,

Xu, Lu, & Chen, 2018), (Ma & Fang, 2020), (Sharma & Batth, 2020), (Yue, Xiaofeng,

& Huagang, 2020), (Williams, 2019), (Abougalala, Amasha, Areed, Alkhalaf, & Khairy,

2020), (Ismail, Hameed, AlShamsi, AlHammady, & Aldhandhani, 2019). Many

researchers also mentioned problems related to the compatibility of blockchain with

existing systems (Raimundo & Rosario, 2021), (Awaji, Solaiman, & Albshri, 2020),

(Sharma & Batth, 2020), (Castro & Au-Yong-Oliveira, 2021), (Liu & Zhu, 2021),

(Pfeiffer, Bezzina, Wernbacher, & Kriglstein, 2020). There is a concern that a battle of

blockchain formats will eventually ensue, resulting in incompatible platforms and

difficulties of data sharing (Gresch & Camilleri, 2017), (Cahyadi, Faturahman, Haryani,

Dolan, & Millah, 2021).

Several authors discussed the adoption barriers arising from the innate characteristics of

blockchain. The immutability feature, for example, was considered detrimental in cases

related to diploma revocation or expiration (Bhaskar, Tiwari, & Joshi, 2020),

(Alammary, Alhazmi, Almasri, & Gillani, 2019), (Raimundo & Rosario, 2021), (Awaji,

Solaiman, & Albshri, 2020), (Loukil, Abed, & Boukadi, 2021), (Yumna, Khan, Ikram, &

Ilyas, 2019), (Castro & Au-Yong-Oliveira, 2021), (Vidal, Gouveia, & Soares, 2019).

Likewise, the transparent decentralised nature of blockchain was seen by some authors

as a threat to privacy of academic information (Alammary, Alhazmi, Almasri, & Gillani,

2019), (Raimundo & Rosario, 2021), (Awaji, Solaiman, & Albshri, 2020), (Ma & Fang,

2020),

(Yumna, Khan, Ikram, & Ilyas, 2019), (Caldarelli & Ellul, 2021), (Castro & Au-

YongOliveira, 2021), (Pfeiffer, Bezzina, Wernbacher, & Kriglstein, 2020), (Ocheja,

Flanagan, Ueda, & Ogata, 2019).

Some scholars also discussed high levels of energy consumption required to run

blockchain (Chen, Xu, Lu, & Chen, 2018), (Gresch & Camilleri, 2017), (Abougalala,

Amasha, Areed, Alkhalaf, & Khairy, 2020). Ghaffar & Hussain (Ghaffar & Hussain,

2019) also mentioned a unique negative aspect called private key loss conundrum which

makes it impossible to recover/change blockchain data if a private key is lost.

Finally, a number of studies pointed to the technology newness, claiming its immaturity

and the lack of both theory and real data for applications in education (Bhaskar, Tiwari,

& Joshi, 2020), (Alammary, Alhazmi, Almasri, & Gillani, 2019), (AlHarthy,

AlShuhaimi, & AlIsmaili, 2019), (Yue, Xiaofeng, & Huagang, 2020), (Williams, 2019),

(Kosmarski, 2020).

Table 12: Technology Barriers to Blockchain Adoption in Education

Barriers Studies

Scalability (Bhaskar, Tiwari, & Joshi, 2020), (Alammary, Alhazmi, Almasri,

& Gillani, 2019), (Awaji, Solaiman, & Albshri, 2020), (Loukil,

Abed, & Boukadi, 2021), (Ma & Fang, 2020), (Sharma & Batth,

2020), (Yue, Xiaofeng, & Huagang, 2020), (Yumna, Khan,

Ikram,

& Ilyas, 2019), (Caldarelli & Ellul, 2021), (Abougalala, Amasha,

Areed, Alkhalaf, & Khairy, 2020), (Ismail, Hameed, AlShamsi,

AlHammady, & Aldhandhani, 2019), (Ocheja, Flanagan, Ueda, &

Ogata, 2019)

Transaction speed over growing data (Chen, Xu, Lu, & Chen, 2018), (Ma & Fang, 2020), (Sharma &

Batth, 2020), (Yue, Xiaofeng, & Huagang, 2020), (Williams,

2019), (Abougalala, Amasha, Areed, Alkhalaf, & Khairy, 2020),

(Ismail, Hameed, AlShamsi, AlHammady, & Aldhandhani, 2019)

Compatibility with legacy systems (Raimundo & Rosario, 2021), (Awaji, Solaiman, & Albshri,

2020), (Sharma & Batth, 2020), (Castro & Au-Yong-Oliveira,

2021), (Liu & Zhu, 2021), (Pfeiffer, Bezzina, Wernbacher, &

Kriglstein, 2020)

Different blockchain platforms /

formats’ clash

(Gresch & Camilleri, 2017), (Cahyadi, Faturahman, Haryani,

Dolan, & Millah, 2021)

Immutability over expired/revoked

credentials (Bhaskar, Tiwari, & Joshi, 2020), (Alammary, Alhazmi, Almasri,

& Gillani, 2019), (Raimundo & Rosario, 2021), (Awaji, Solaiman,

& Albshri, 2020), (Loukil, Abed, & Boukadi, 2021), (Yumna,

Khan, Ikram, & Ilyas, 2019), (Castro & Au-Yong-Oliveira,

2021), (Vidal, Gouveia, & Soares, 2019).

Privacy over transparent nature of

blockchain

(Alammary, Alhazmi, Almasri, & Gillani, 2019), (Raimundo &

Rosario, 2021), (Awaji, Solaiman, & Albshri, 2020), (Ma &

Fang, 2020), (Yumna, Khan, Ikram, & Ilyas, 2019), (Caldarelli &

Ellul, 2021), (Castro & Au-Yong-Oliveira, 2021), (Pfeiffer,

Bezzina, Wernbacher, & Kriglstein, 2020), (Ocheja, Flanagan,

Ueda, & Ogata, 2019)

Technology newness (Bhaskar, Tiwari, & Joshi, 2020), (Alammary, Alhazmi, Almasri,

& Gillani, 2019), (AlHarthy, AlShuhaimi, & AlIsmaili, 2019),

(Yue, Xiaofeng, & Huagang, 2020), (Williams, 2019),

(Kosmarski, 2020)

High level of energy consumption (Chen, Xu, Lu, & Chen, 2018), (Gresch & Camilleri, 2017),

(Abougalala, Amasha, Areed, Alkhalaf, & Khairy, 2020)

Private key loss conundrum (Ghaffar & Hussain, 2019)

The identified organisational barriers to blockchain adoption in education are related to

costs of implementation, insufficient knowledge, and possible resistance to change

(Table 13). Researchers noted the high level of out-of-pocket costs for blockchain

implementation as well as supplemental costs for system maintenance (Alammary,

Alhazmi, Almasri, & Gillani, 2019), (Awaji, Solaiman, & Albshri, 2020), (Yue,

Xiaofeng, & Huagang, 2020) and increasingly high resource use (Gresch & Camilleri,

2017). Potential issues with financing and infrastructure were also discussed by

Cahyiadi et al. (Cahyadi, Faturahman, Haryani, Dolan, & Millah, 2021). Likewise, a

number of researchers pointed to a general lack of understanding of blockchain by both

educators and school administrators (Yue, Xiaofeng, & Huagang, 2020), (Williams,

2019), (Pfeiffer, Bezzina, Wernbacher, & Kriglstein, 2020), (Abougalala, Amasha,

Areed, Alkhalaf, & Khairy, 2020). Ma and Fang (Ma & Fang, 2020) also mentioned that

for many education institutions, the implementation of blockchain is difficult because of

lack of skilled specialists in this area whereas Fedorova and Skobleva (Fedorova &

Skobleva, 2020) noted a general lack of awareness of blockchain and its benefits among

decision makers in the educational sector. In addition, blockchain is often perceived as

too complex (Castro & Au-Yong-Oliveira, 2021), (Liu & Zhu, 2021), (Gresch,

Rodrigues, Scheid, Kanhere, & Stiller, 2018), bringing new dependencies on third

parties (Gresch, Rodrigues, Scheid, Kanhere, & Stiller, 2018), and blurring property

rights (Ma & Fang, 2020). Finally, some researchers were concerned about the possible

conflict of values for those seeking to preserve traditional education approaches

(Alammary, Alhazmi, Almasri, & Gillani, 2019), (Fedorova & Skobleva, 2020),

(Kosmarski, 2020). These individuals may be resistant to blockchain adoption.

Table 13: Organisational Barriers to Blockchain Adoption in Education

Barriers Studies

Initial costs and maintenance costs (Alammary, Alhazmi, Almasri, & Gillani, 2019),

(Awaji, Solaiman, & Albshri, 2020), (Yue,

Xiaofeng, & Huagang, 2020), (Gresch & Camilleri,

2017), (Cahyadi, Faturahman, Haryani, Dolan, &

Millah, 2021)

Finally, environment adoption barriers (Table 14) are mostly viewed through the prism

of the lack of a regulatory environment (Fedorova & Skobleva, 2020) and the lack of

clarity regarding compliance with privacy laws (Vidal, Gouveia, & Soares, 2019). This

may result in possible legal issues for education institutions (Loukil, Abed, & Boukadi,

2021), (Abougalala, Amasha, Areed, Alkhalaf, & Khairy, 2020), (Kosmarski, 2020).

Cahyadi et al. (Cahyadi, Faturahman, Haryani, Dolan, & Millah, 2021) also pointed to

the lack of standardization in blockchain technology for education, while Ma and Fang

(Ma & Fang, 2020) discussed the lack of general policy guidance and protection

mechanisms.

Table 14: Environmental Barriers to Blockchain Adoption in Education

2.7 Adoption Factors

Factors influencing the adoption of blockchain by education institutions, unlike the

barriers hitherto, remain a relatively unexplored area. Only 8 studies in the sample

discussed factors that would positively influence blockchain adoption in education.

Institutional cooperation was discussed in 5 studies (Eaganathan et al., 2019; Gresch &

Camilleri, 2017; Lizcano et al., 2020; Ocheja et al., 2019; Widjaja et al., 2020). Gresch

& Camilieri (2017) also mentioned that state-level cooperation on the standardization of

blockchain mechanisms should influence the speed of adoption. Widjaja et al. (2020)

and El Nokiti and Yusof (2019) pointed to the need of raising blockchain awareness and

knowledge for application in education. Yue et al. (2020) proposed that influential

blockchain adoption mechanisms would include the development of safe and reliable

rules of use, creating an efficient technology transition path, establishing mechanisms

for open data sharing, and reforming the education management system. However,

Perceived complexity (Yue, Xiaofeng, & Huagang, 2020), (Williams,

2019), (Pfeiffer, Bezzina, Wernbacher, & Kriglstein,

2020), (Abougalala, Amasha, Areed, Alkhalaf, &

Khairy, 2020)

Conflict of values (Alammary, Alhazmi, Almasri, & Gillani, 2019),

(Fedorova & Skobleva, 2020), (Kosmarski, 2020)

Lack of skilled personnel (Ma & Fang, 2020)

Lack of technology awareness (Fedorova & Skobleva, 2020)

Dependencies on third parties (Gresch & Camilleri, 2017)

Blurred property rights (Ma & Fang, 2020)

Ullah et al. (2020) conducted the only empirical study to examine the influence of

adoption factors. Trialability, compatibility, and relative advantage were confirmed as

positive influencers of blockchain adoption in education both directly and via perceived

usefulness and perceived ease of use.

Therefore, it can be concluded that there is a clear lack of empirical research on the

factors influencing the adoption of blockchain in education. Ullah’s study was a

welcoming addition to the existing knowledge, although it only offered an empirical

investigation of blockchain adoption factors at an individual level. For the most part, it

is education administrators who make decisions related to new technology trials and

adoption in industry. Therefore, more empirical research on organisational factors is

required. This represents an important area for investigations in the future, especially

given the relatively slow pace of the adoption of blockchain in this field.

2.8 Theoretical Frameworks of Adoption

While multiple conceptual frameworks for blockchain implementation have been

proposed, only one study in the sample applied an established theoretical framework for

blockchain adoption in education. Ullah et al. (2020) used an integrated Technology

Acceptance – Diffusion of Innovations (TAM-DOI) framework to examine blockchain

acceptance in smart learning environments. However, the framework only examined

individual-level factors related to how blockchain is perceived by users. The lack of

theoretical grounding for blockchain adoption in education can be explained by its

relative immaturity.

To some extent, the adoption of blockchain in education could be examined through the

theoretical lens of general blockchain adoption frameworks developed with no specific

link to a particular industry (Table 15). The Technology-Organization-Environment

(TOE) framework and its expansions are the most popular in this regard, although some

authors also proposed to explore adoption factors through the lens of the Theory of

Reasoned Action (TRA) and a value driver perspective. It is difficult to predict the

explanatory power of these frameworks since they do not consider education

environment specifics. Overall, theory development for blockchain adoption in

education offers a relatively unexplored but potentially rather fertile ground for further

research.

Table 15: General and Education Industry Specific Blockchain Adoption Frameworks

Study Framework

Type

Theory or

Perspective Used

Factors Considered

Angelis & Da

Silva (2019)

General Value driver

perspective

Value opportunities (9 factors), value drivers (4

factors), technology feasibility and viability (7

factors)

Barnes & Xiao

(2019)

General TOE Technological (5 factors), organisational (4

factors), environmental (5 factors)

Clohessy et al.

(2020)

General TOE enhanced Technological (15 factors), organisational (13

factors), environmental (12 factors), task-

related (12 factors), individual (13 factors)

Janssen et al.

(2020)

General TOE variation Institutional (3 factors), market (3 factors),

technology (3 factors)

Li (2020) TRA-TAM Perceived usefulness, perceived ease of use,

perceived trend, subjective norms, attitude to

adopt, intention to adopt

Toufaily et al.

(2021)

General TOE Technological (8 factors), organisational (3

factors), environmental (3 factors)

Ullah et

al. (2020)

Specific TAM-DoI Trialability, relative advantage, compatibility,

ease of use, perceived usefulness

2.9 Search Limitations

The study results should be considered within several limitations. One possible

limitation is the number of databases selected for the extraction of the studies. On the

one hand, it is physically difficult to review all possible sources of literature online in a

limited amount of time. On the other hand, we believe that the selected databases, due to

their reputation and scale, provide a good collection of research on the topic which is

both illustrative and encompassing. This study also used an established, practical

approach to filtering the literature which helped resolve possible literature selection

biases. Another limitation is the rapid nature of change that blockchain may bring to the

education industry. As such, the study findings may soon require updating. However, the

usefulness of this research may also be seen by comparing the state of blockchain

adoption at this and later points in time. This would allow the progress of technology

adoption to be tracked, as well as the practical and theoretical developments that it

brings. Finally, as is the case with many reviews, this study offers conclusions based on

the researchers’ subjective opinions. However, the study did not aim to measure the

actual adoption of blockchain in education institutions or the state of such adoption.

Instead, it aimed to organise the existing knowledge on the topic and point to the gaps in

this knowledge. In this regard, the findings of the study are rather useful in guiding

empirical studies in the future.

2.10 Research Gaps

The systematic literature review in Chapter 2 highlighted some important gaps that exist

in the literature on blockchain adoption in higher education:

1. The studies on blockchain technology adoption in education are limited. Studies

in the education sector represent a fraction of studies on blockchain adoption in

other sectors like finance, supply chains and logistics and healthcare. This

demonstrates the need to expand the body of knowledge on blockchain adoption

in the education sector.

2. The majority of knowledge on blockchain adoption so far has been formed

through review studies and conceptual models. They cover either existing or

potential uses of blockchain without paying particular attention to combinations

of factors that promote or impede it. This demonstrates the need to explore such

factors.

3. There are very few studies that present theoretically well-founded models of

blockchain adoption in education. An analysis of the relevant literature suggests

the absence of a theoretical-based study that serves as a guide in developed and

developing countries to explain blockchain adoption in higher education. This

demonstrates the need to further explore the relevant adoption theories to better

describe the process, present and empirically test such theoretical frameworks.

4. There is a dearth of qualitative research on the topic of blockchain adoption in

higher education. However, qualitative research is very useful in describing

blockchain perceptions of education administrators, users (students/instructors),

and technology experts. These can be compared to the actual state of technology

in educational institutions that successfully adopted and used it.

5. To the best knowledge of the author, there is no current work on adopting

blockchain technology in higher education performed in Arab countries,

particularly in KSA. Saudi researchers so far have focused almost exclusively on

reviews of blockchain adoption cases abroad. This demonstrates the need to

conduct blockchain adoption research in the Saudi context.

From the gaps identified in the existing literature on blockchain adoption in higher

education, it follows that there is a need to explore the factors that promote and impede

the adoption process. Further, it follows that such factors need some clear dimensional

classification and grouping for ease of comprehension, review, and actionability. The

existing studies that applied adoption frameworks were analysed along the following

dimensions: technology, organisational, environmental, and quality/value (Table 16).

Therefore, this study considers these four dimensions as influential in the adoption

process.

Table 16: Factor Dimensions in Blockchain Adoption Studies

Study Framework Technology

Factor Dimensions

Organisation

Environment Quality

Angelis & Da Silva

(2019) Value drivers x x

Barnes & Xiao (2019) TOE x x x

Clohessy et al. (2020) TOE x x x x

Janssen et al. (2020) TOE x x x

Li (2020) TRA x

Toufaily et al. (2021) TOE x x x

Ullah et al. (2020) TAM-DoI x x

2.11 Chapter Summary

This chapter sought to provide an up-to-date relevant picture of the contemporary status

of research on blockchain adoption in education. Based on the review of 107 studies, a

taxonomy of the research was developed which offers an organised, ordered view on the

state of research – something which was lacking in the previous literature reviews on the

topic. The taxonomy contributes to the research field by offering a roadmap for future

studies based on easily identifiable gaps in research and knowledge.

Overall, the reviewed literature demonstrated the expanding research on blockchain

adoption in education. However, the majority of knowledge on blockchain adoption so

far has been formed through review studies and conceptual models. While they offer

important insights and present potentially viable technical solutions, they do not provide

useful information on what factors contribute to the adoption process. Nevertheless, a

knowledge of such factors can benefit organisational decision makers who may consider

blockchain applications but remain unsure of their usefulness and potential for their

institutions. Perhaps the lack of practical information on blockchain adoption is the

reason why many administrators lack motivation for blockchain adoption and prefer a

wait-andsee strategy (Ma & Fang, 2020). Therefore, it is necessary for researchers to

provide more data on actual adoption cases and explore the factors contributing and/or

impeding the adoption process.

Several research avenues arise from here. First, simple reviews of use cases may not be

sufficient to offer a good practical view of the adoption process. Case studies of

successful blockchain adoption as well as action research could provide practical

insights into the process and its outcomes and serve as a viable approach to transfer

actual industry experience to academic research. This, in turn, will enable a systematic

view of the adoption process to be formed and avoid trial-and-error approaches to

adoption in educational settings. Second, this review identified only 5 empirical

investigations of blockchain adoption, and only one of them was quantitative research.

Obviously, this is very insufficient for such a dynamically developing field. Qualitative

research can be useful in this regard by investigating blockchain perceptions from

education administrators, users (students/instructors), and technology experts. These can

be compared to the actual state of technology in educational institutions that

successfully adopted and used it. Further, the field clearly lacks a strong theoretical

grounding for the adoption process. This is in deep contrast to, for example, supply

chain management where several adoption theories have been successfully introduced

and tested (Alazab et al., 2021; Choi et al., 2020; Kouhizadeh et al., 2021; Queiroz &

Wamba, 2019).

Producing a strong theoretical framework for blockchain adoption in higher education

will serve as a good foundation for empirical research in the field. One major gap in the

knowledge in this regard is the absence of a holistic blockchain adoption framework

which is empirically tested to identify the effect of various types of factors (technology,

organisational, environmental and quality). To the best knowledge of the author, there is

no such framework for the Saudi higher education context. The next chapter defines the

research problem based on the identified research gaps and outlines the key research

question and subquestions.

3 Problem Definition

3.1 Introduction

This chapter presents the problem definition that guides the research process in this

study. Section 3.2 outlines the key definitions of the terms used in the chapter. Section

3.3 outlines the key research questions to be addressed in the study. Section 3.4 presents

the main study objectives guided by the research problem and questions.

3.2 Key Definitions

Given below is the list of the key terms used in this chapter.

Blockchain: a decentralized, digital ledger that records transactions in a secure,

transparent and permanent manner. It operates through a network of computers, where

each block in the chain contains a list of transactions that are verified and added to the

chain through cryptography. This makes the data stored in a blockchain immutable and

resistant to tampering, creating a trustless system without the need for intermediaries.

Blockchain technology is primarily used in the context of cryptocurrencies, but it has

potential applications in various fields including higher education (Alammary, Alhazmi,

Almasri, & Gillani, 2019).

Higher Education Institution (HEI): an organisation that offers post-secondary

education leading to a degree or professional qualification. These institutions typically

provide programs of study at the undergraduate and graduate level, and usually include

universities, colleges, community colleges and technical schools (Alenezi, 2021).

Holistic Framework: a comprehensive approach to problem solving or decision making

that takes into account all relevant aspects and factors, rather than just a single isolated

aspect. This type of framework aims to provide a complete and integrated perspective,

considering the interconnections and relationships between various elements, to arrive at

a well-rounded solution or conclusion.

Technological Factors: features of a specific technology that define it and may

influence its adoption and use. This study looks into a set of technology factors outlined

in the Diffusion of Innovation theory by Rogers (1995).

Organisational Factors: the internal characteristics of an organisation that can affect its

operation and performance, the way an organisation functions, its ability to adapt to

change, and its ability to achieve its goals and objectives. In the context of technology

adoption, organisational factors play a critical role in determining whether a technology

in question represents a good fit with the organisation. This study considers the

organisational factors outlined by Tornatzky et al. (1990).

Environmental Factors: the institutions and processes that create the broad context of

organisational operations. While they exert a certain influence on organisations, the

organisations do not have the power to influence them in return. Some examples are

market conditions, competitive action, national culture and political environment. This

study considers the organisational factors outlined by Tornatzky et al. (1990).

Quality Factors: the elements or attributes that determine the degree of excellence or

merit of a product, service, or experience. These factors can vary depending on the

industry, customer preferences, and the type of product or service being evaluated. In

the context of higher education, quality factors play a critical role in determining

students’ satisfaction and can impact their competitiveness in the labour market (Harvey,

2007).

Barriers to adoption: the obstacles that prevent organisations from fully embracing and

utilising new technology. These barriers can range from practical considerations such as

cost, compatibility with existing systems, and ease of use, to more intangible factors

such as resistance to change, lack of understanding, or cultural attitudes (Clohessy,

Treiblmaier, Acton, & Rogers, 2020). Overcoming these barriers can be critical for

organisations looking to remain competitive in an increasingly technology-driven world.

3.3 Research Questions

Based on the gaps identified in the literature, the main research question in this study is:

How can we develop a holistic blockchain adoption model for Saudi HEIs?

As previously discussed, there is no comprehensive model for blockchain adoption by

HEIs in Saudi Arabia’s context. Further, even though some attempts to develop such

frameworks have been undertaken in other national contexts (see Table 15 in Section

2.10), the quantity and type of factors included in those models vary from study to

study. This suggests that even though the grouping across dimensions remains more or

less consistent, different factors would be included based on context. This study

attempts to identify the specific factors in the context of Saudi Arabia HEIs.

Accordingly, the following research sub-questions are formulated:

RQ1: Which technological factors influence Saudi higher education institutions’

intention to adopt blockchain technology?

RQ2: Which organisational factors influence Saudi higher education institutions’

intention to adopt blockchain technology?

RQ3: Which environmental factors influence Saudi higher education institutions’

intention to adopt blockchain technology?

RQ4: Which quality factors influence Saudi higher education institutions’ intention to

adopt blockchain technology?

In addition to the factors that enable blockchain adoption in higher education settings,

there are also a number of barriers that prevent or slow it down. As discussed in Section

2.6 of this thesis, these barriers can be categorised as: technology related, organisation

related, and environment related. Again, however, there is no consistency in the types or

number of barriers that pertain to each of the aforementioned dimensions. Therefore, the

following research question and sub-questions are formulated:

RQ5: What are the barriers to adopting blockchain technology in Saudi higher

education institutions?

-What are the technology barriers?

-What are the organisational barriers?

-What are the environmental barriers?

3.4 Research Objectives

Based on the main research question in this study, the primary objective is:

To develop a holistic blockchain adoption framework that can be practically applied

by Saudi HEIs.

Given the gaps identified in the literature, this study aims to develop and validate a

comprehensive adoption model for blockchain technology in the context of higher

education in Saudi Arabia. While adoption models are already available in the literature,

this research follows the major view that factors of adoption will vary depending on

context. Initial qualitative research would help identify specific factors that could be

relevant for the considered context. These, in turn, could be tested on a large population

to determine the strength and direction of their relationship with adoption intent. To

effectively meet the key objective formulated in this study, it is necessary to investigate

the factors that influence (positively or negatively) the adoption of blockchain

technology in Saudi HEIs across several dimensions. Accordingly, the accompanying

objectives for the research are:

•Develop an initial comprehensive model of blockchain technology adoption in

Saudi higher education institutes based on the relevant theoretical foundations

and empirical literature;

This objective is met by reviewing the relevant technology adoption theories and

selecting the relevant factors whose influence has been confirmed empirically in the

studies on blockchain adoption in education.

•Refine the model to fit with the context of Saudi higher education institutions;

This objective will be met upon completion of Phase I of the research which involves

interviews with the decision makers in Saudi HEIs. The analysis of the interviews will

allow the factors found non-influential to be excluded and some new factors to be

added, possibly not explored in the previous studies but found relevant in the context of

Saudi HEIs. It is expected that some factors could be revised or combined for the same

reasons.

•Test the relationships within the finalised framework empirically to identify the

factors influential in the blockchain adoption process by Saudi HEIs;

This objective will be met upon completion of Phase II of the research which involves

quantitative tests of the framework relationships. The analyses will show: 1) whether the

model itself is fitting in exploring the adoption factors; and 2) what factors are

influential in the adoption process within the selected context.

•Based on the study findings, offer new insights of both theoretical and practical

kinds to blockchain adoption and applications in higher education.

This objective will be met by placing the study results within the existing body of

knowledge on blockchain adoption in education. The results of the research will be

compared to those reported in the literature and to the hypotheses formulated in this

study. Particular attention will be given to discrepancies in the findings against

expectations.

Contributions to theory and practice will be reported.

3.5 Chapter Summary

This chapter introduced the research problem addressed in this study, presented the

research question and formulated the key research objective. These were developed

based on the gaps in the knowledge and research approaches identified in the course of

the literature review. The key research objective posed in this study is to develop a

practical holistic blockchain adoption framework for Saudi HEIs. Five research

subquestions and corresponding objectives were formulated and explained. The next

chapter provides an overview of the solution to the research problem based on the

literature review of the relevant theories and approaches.

4 Solution Overview and Methodology

4.1 Introduction

This chapter presents a solution overview to the formulated research question and

outlines the approach to implement it. Section 4.2 presents the definitions of the key

technical terms used in the chapter. Section 4.3 describes the overall solution approach

and the selection of the theories to form the foundation of the model for blockchain

adoption in Saudi HEIs. Section 4.4 reviews the solutions aligning with research

subquestions 1-5. Section 4.5 presents the research methodology and justifies a mixed

methods design to test and validate the model. Section 4.6 details the process of model

presentation and the initial analysis by a group of industry experts. Section 4.7 describes

in detail the approach for model evaluation on a large sample of professionals from the

Saudi higher education sector. Finally, Section 4.8 outlines the key procedures to ensure

high ethical standards for the data collection and analysis.

Sections of this chapter have earlier been published in the following conference article:

Alalyan, M.S., Jaafari, N.A., Hussain, F.K. (2023). Technology factors influencing

Saudi higher education institutions’ adoption of blockchain technology: A

qualitative study. Advanced Information Networking and Applications (AINA).

4.2 Key Definitions

Presented below are the definitions of the key technical terms used in this chapter.

Theory: a set of interrelated concepts, definitions, and propositions that provide a

systematic explanation of a phenomenon. Theories are developed based on empirical

evidence and logical reasoning, and they are designed to be testable/falsifiable. The

choice of a theory is important to contextualize and make sense of the findings,

contributes to the advancement of knowledge, and improves the validity and reliability

of the results (Creswell J. W., 2014).

Research Framework: the conceptual structure or blueprint that guides a research

study. It includes the theoretical foundations, research questions, hypotheses, variables,

methods, and data analysis techniques that shape the design and implementation of the

study. A research framework provides a roadmap for conducting the research and helps

ensure that the study is logically consistent, coherent and well-organised (Bryman,

2016).

Research Design: plan or strategy for conducting a study that involves collecting and

analysing data to answer a specific research question. It outlines the methods,

techniques, and procedures that will be used to collect and analyse data, and it provides

a framework for conducting the study in a systematic, rigorous, and credible manner.

The research design is an important component of the research process as it helps to

ensure that the study is well-designed, appropriately executed, and that the results are

reliable and valid (Creswell & Creswell, 2017).

Research Methodology: the specific procedures and techniques used in conducting a

research study. It includes the design, data collection, data analysis, and interpretation of

results, among other aspects. A research methodology provides a framework for

carrying out a systematic and organised investigation and helps ensure that the study is

rigorous and credible (Creswell J. , 2015).

Qualitative Research: a type of research that seeks to understand human behaviour and

the reasons behind it. It focuses on the subjective experiences and perceptions of

individuals, and often involves in-depth and open-ended data collection methods, such

as interviews, focus groups, and observation. The aim of qualitative research is to

provide a rich, detailed, and nuanced understanding of the experiences and perspectives

of the participants, rather than to test a specific hypothesis or generate numerical data

(Olfazoglu, 2017).

Quantitative Research: a type of research that uses numerical and statistical data to

understand and describe phenomena. It involves the collection and analysis of

quantitative data, through standardised and structured methods such as surveys,

experiments or observations. The aim of quantitative research is to test hypotheses,

identify relationships and patterns, and make generalisations about a larger population

based on the results of the study. Quantitative research often uses statistical techniques

to analyse the data and make inferences (Creswell & Creswell, 2017), and it is used in

fields such as economics, psychology, sociology, and public health, among others, to

provide a numerical understanding of the phenomena being studied.

Mixed Research: a type of research that combines both qualitative and quantitative

research methods in a single study. This approach allows for the examination of a

phenomenon from multiple perspectives and provides a more comprehensive

this, a key research question was formulated (Chapter 3). A granular view of the

research question was provided given the complexity of the issue at hand and

subquestions were formulated accordingly (Chapter 3). The value of the solution was

presented initially in Chapter 1 where the role of novel technologies in higher education

was discussed.

Step 2: define the solution objectives. In this step, the key objective was formulated on

the basis of the main research question. The goal of the solution is to provide an

actionable framework of blockchain adoption for Saudi HEI administrators. Several

related objectives were formulated in Chapter 3 for a holistic approach to the issue.

Resources to meet the objectives were discussed as well: available theoretical and

empirical studies of adoption, industry experts’ opinions, and a large-scale survey to

confirm the role of specific model elements.

Step 3: design and development. In this step, the initial solution is created, which is an

original blockchain adoption framework for Saudi HEIs. The development of the initial

model is based on: 1) review and evaluation of the available theories and frameworks;

2) integration of the suitable theories; and 3) identification of the key model elements on

the basis of theoretical and empirical studies. This process is thoroughly described in the

remaining parts of Chapter 4. Chapter 5 then describes the original, conceptualised

blockchain adoption framework and explains the relationships among its elements.

Step 4: demonstration. In this step, the initial model is demonstrated to a group of

industry experts for review and evaluation. This constitutes Phase I of the research

process, where the experts formulate professional opinions on the framework and its

elements as well as propose additional elements which they deem appropriate for the

context of Saudi HEIs. The key outcome of this step is a refined framework of

blockchain adoption by Saudi HEIs. This step is undertaken in Chapter 6.

Step 5: evaluation. In this step, the refined model is tested on a large group of

education and IT professionals from Saudi HEIs. This represents Phase II of the

research process. The outcome of this step is the empirical evidence of how well the

model explains blockchain adoption in Saudi HEIs as well as the efficacy of each

predicted model element in this process. This step is detailed in Chapter 7.

Step 6: communication. The final step consists of summarising the research findings,

analysing them and presenting them to the relevant audiences. A discussion of the

research outcomes and conclusions is provided in Chapters 8 and 9 respectively. The

literature review results have been communicated through a publication in a

peerreviewed journal. Additional publications are expected after the completion of the

thesis.

As previously explained, the solution approach begins with the choice of the best fitting

adoption theories which will form the foundation of the adoption model.

4.3.2 Theoretical Framework Choice

The positive impact of IT innovations on nearly every element of human life and

organisational performance is well recognised and reported in the literature (Clohessy &

Acton, 2019; O'Connor, Lowry, & Treiblmaier, 2020). The changes brought by

innovative technologies became especially visible in the past few decades with the rapid

development and integration of computers, the internet, mobile devices and

telecommunications. This, in turn, increased the demand for theories and frameworks

that could help understand how and why technological innovations are adopted and

what factors play significant roles in the process (Khan & Qudrat-Ullah, 2020; Lai,

2017).

A number of theories and frameworks have been proposed to explain the phenomenon

of adopting technology innovations. They focus on different underlying mechanisms of

adoption which, in turn, leads to different factors considered influential. For this reason,

it is necessary to have a comprehensive understanding of the popular models and

theories, their key assumptions as well as the strengths and limitations to identify the

one which fits the purpose and the context of this study best. Therefore, the following

sections provide a critical review of several widely used technology adoption theories

and frameworks. The rejected theories/models are reviewed first with an explanation of

why they were deemed unfit for the study. Finally, the choice of the Diffusion of

Innovations (Rogers, Diffusion of innovations, 1995) theory and Technology-

OrganisationEnvironment (Tornatzky & Fleischer, 1990) framework as the foundations

for this study is justified.

4.3.2.1 Social Cognitive Theory

Social Cognitive Theory (SCT) takes root in social psychology studies that focused on

human actions and reactions to particular phenomena (Bandura, 1986). The theory

views the adoption of innovation as an interplay of personal characteristics, behaviour

and environment (Figure 10). Individual behaviours interact with personality through

thoughts and actions; personal factors interact with the environment through the

formation of beliefs and competencies; and behaviour interacts with the environment

through the understanding and modification of actions.

Figure 10: Social Cognitive Theory (Bandura, 1986)

SCT makes the proposition that individuals interact with the environment and their peers

and receive feedback through which they learn and modify their further actions (Tarhini,

Arachchilage, Masa'deh, & Abbasi, 2015). Bandura (1986) described it the following

way:

Social cognitive theory distinguishes among three separable components

in the social diffusion of innovation. The triadic model includes the

determinants and mechanisms governing the acquisition of knowledge

and skills concerning the innovation; adoption of that innovation in

practice; and the social network by which innovations are promulgated

and supported. (p. 119)

Technology adoption in SCT is a type of social process and a product of gradual

understanding of the innovation in question, acquiring skills for its use, and receiving

positive feedback from personal networks and the environment. SCT is helpful in

predicting adoption through individual and group behaviour change. It also takes into

account the environment, which is important for studying innovations in different

contexts. Yet, this theory is strongly positioned towards learning and personal actions,

which is hardly surprising given its social study roots. On the other hand, it is not

wellpositioned with respect to technology because it ignores the effect of any potential

technology characteristics. Further, SCT is focused on individuals, which makes it less

appropriate for organisational settings which is the focus of this study. For these reasons,

SCT was not considered as a theoretical foundation for this research.

4.3.2.2 Theory of Reasoned Action

Behaviour

Personal

Factors

Environment

Factors

Beliefs,

competencies

Thoughts,

actions

Understanding,

Modification

Theory of Reasoned Action (TRA) was also originally developed for sociological and

psychological research and later applied for information systems adoption studies

(Taherdoost, 2018). The focus of TRA is on human behaviour which is seen as a product

of rational decision making where implications of actions are considered (Ajzen &

Fishbein, Understanding attitudes and predicting social behaviour, 1980). The process of

forming particular behaviours as outlined by TRA is presented in Figure 11. Behaviour

is represented by concrete actions that arise from behavioural intentions – a plan to

engage in those actions. Behavioural intentions, in turn, are products of a combination

of attitudes and subjective norms. Fishbein and Ajzen (1975) defined attitudes as the

person’s own evaluation of performing the intended action and subjective norms as

perceptions of what close people would think of it. Accordingly, attitudes arise from

personal beliefs while subjective norms from prevailing normative (social) beliefs.

Figure 11: Theory of Reasoned Action (Ajzen & Fishbein, 1980)

From the perspective of TRA, technology adoption is the behavioural endpoint of the

framework presented in Figure 11. The intent to use it would result from what

individuals think about the technology themselves and what they believe others think

about its use. Positive personal attitudes and subjective norms encourage intention to

use. TRA can be considered more comprehensive than SCT in terms of internal drives to

adopt an innovation. However, this comes at the expense of environmental factors since

they are omitted from the framework. Environmental factors, as discussed above, are

essential in considering technology adoption in different contexts. Further, as Tarhini et

al. (2015) argued, TRA is a general model: it cannot be effectively adjusted to specific

behaviours such as adoption of certain, non-conventional types of technologies. The

model seems to be more appropriate for examining adoption of the already established

technologies for which subjective norms and attitudes have been formed, which is not

the case with emerging technologies like blockchain. Finally, like SCT, TRA focuses on

individuals and ignores technological factors that could affect attitudes and, hence,

intention to use technology. For these reasons, TRA was not considered as a theoretical

foundation for the current study.

4.3.2.3 Theory of Planned Behaviour

Theory of Planned Behaviour (TPB) can be considered a modification of TRA: it was

developed by the same researchers and introduced a single new element to the

framework. According to Ajzen (1991), the purpose of the theory was to consider the

absence of a complete volitional control over one’s behaviour. To do this, TPB

introduced a variable of Perceived Behaviour Control which is ‘perceived ease or

difficulty of performing the behaviour’ (Ajzen, 1991, p. 188). This variable considers

internal as well as external behavioural constraints that could impede action even in the

presence of the right attitudes and subjective norms (Taylor & Todd, Decomposition and

crossover effects in the theory of planned behavior: A study of consumer adoption

intentions, 1995). In other aspects, TPB preserves the logic of TRA by considering

attitudes and subjective norms as leading to behavioural intention and to actual

behaviour in the end (Figure 12).

Figure 12: Theory of Planned Behaviour (Ajzen, 1991)

In the field of technology innovations, TPB not only considers personal attitudes and

subjective norms in the adoption process, but also pays attention to what could limit

one’s behaviour. For example, a potential technology user would consider whether she

has the required access and resources for technology use and meet the requirement

criteria.

By introducing the notion of behavioural controls in the form of additional internal and

external factors, TPB advanced TRA. It is also considered a more flexible model, and

several extensions and modifications have actually been introduced to fit the needs of

specific researchers (Conner & Armitage, 1998; Taylor & Todd, Decomposition and

crossover effects in the theory of planned behavior: A study of consumer adoption

intentions, 1995). Still, like its predecessor TRA, TPB concentrates on an individual

user, their characteristics and behaviours while not considering technology factors or

organisational characteristics. Further, even with the introduction of the control variable,

the link between behavioural intention and actual behaviour has not been clearly

established for this framework (Tarhini et al., 2015). Therefore, TPB was also rejected

as a theoretical foundation for this study.

4.3.2.4 Task-Technology Fit

The TTF model (Figure 13) was developed by Goodhue and Thompson (1995) in

response to the dominance of behaviour-oriented theories in information systems

studies. The researchers proposed that a technology would be utilised as a result of the

congruence of its characteristics with the tasks that individuals want to perform with it

(Lai, 2017). The task-technology fit is a comprehensive measure that consists of eight

factors: quality, locatability, authorisation, compatibility, ease of use/training, production

timeliness, systems reliability and relationship with users (Goodhue & Thompson,

1995). Accordingly, a good task-technology fit is considered an indicator of improved

task performance by technology users. Within TTF, users adopt blockchain because its

inherent characteristics match the issues that users need to resolve: for example,

blockchain’s immutability feature matches the task of ensuring college certificate

legitimacy and security.

Figure 13: Task-Technology Fit Model (Goodhue & Thompson, 1995)

Unlike the behaviour-oriented theories reviewed above, TTF is a model that is: 1) more

technology oriented and 2) considers tasks at hand as an important component of

adoption. Furthermore, even though the original TTF was directed at individual users,

its later adaptations also focused on group-level performances such as potential

influence of technology on effectiveness and efficiency, quality of output and overall

group satisfaction (Delgado Piña, María Romero Martínez, & Gómez Martínez, 2008;

Zigurs & Buckland, 1998). This makes it potentially more adaptable for research at the

organisational level. However, due to the focus on utilisation and performance, the

model could be better suited to test for the outcomes of adoption rather than the process

of adoption. Further, since the model assigns much importance to the task, it could be

better suited to explore technology adoption for a specific organisational function rather

than a comprehensive, organisation-wide project. While bringing technology factors into

the limelight, TTF clearly ignores organisational and environment characteristics.

Finally, it should be noted that empirical studies often provide inconsistent results for

different technologies and contexts which makes the model difficult to generalise

(Daradkeh, 2019; Eybers, Gerber, Bork, & Karagiannis, 2019; Lin, Wu, Lim, Han, &

Chen, 2019). For these reasons, TTF was not considered as a foundation for the current

research.

4.3.2.5 Technology Acceptance Model and Its Variations

Technology Acceptance Model (TAM) arose from TRA, but it was developed with a

specific focus on technology. The original idea for the model was proposed by Davis

(1989) who stated that behavioural intention to use a technology arises from specific

beliefs regarding the technology rather than generic attitudes. These beliefs were

summarised within two key variables: perceived ease of use (PEOU) and perceived

usefulness (PU) (Davis, 1993). The refined TAM envisioned behavioural acceptance of

technology as a sequential process (Figure 14). According to the model, technology use

arises from the intention to use, which is influenced by positive attitudes towards the

technology. These, in turn, are predicted by whether the potential users believe that the

technology is easy to use (PEOU) and beneficial (PU). The model further proposes that

if a technology is perceived easy to use, it also influences the degree of its perceived

usefulness. Finally, the perceived technology variables in the model are influenced by

external factors which could be either technology related (such as system design,

available documentation, technical support) or user related (such as enjoyment or

training) (Agarwal & Karahanna, 2000; Davis, Bagozzi, & Warshaw, 1989). Earlier

studies using TAM, however, showed that both PEOU and PU had a consistent direct

effect on behavioural intention, which allowed the model to be simplified by eliminating

the attitude variable (Venkatesh & Davis, A model of the antecedents of perceived ease

of use: Development and test., 1996).

Figure 14: First Finalised TAM (Davis et al., 1989)

The development of TAM has contributed significantly to the analysis of user

motivations in technology acceptance and actual use. The model has been validated

multiple times across different types of technologies with excellent predictive results

(Khan & QudratUllah, 2020; Taherdoost, 2018; Tarhini, Arachchilage, Masa'deh, &

Abbasi, 2015). TAM also proved rather flexible in nature, as several extensions and

modifications with various degrees of complexity have been introduced: A-TAM (Taylor

& Todd, 1995), TAM2 (Venkatesh & Davis, 2000), UTAUT (Venkatesh, Morris, B, &

Davis, 2003) and TAM3 (Venkatesh & Bala, 2008) being the most commonly mentioned

and explored in the literature. The later models offered a very granulated view of the

external factors that contribute to increased PEOU and PU. For example, TAM3 contains

13 factors either directly or indirectly influencing these two variables.

The contribution of TAM and its several versions to technology-related studies has been

enormous. The model and its modifications is still being actively used by researchers in

various fields, including studies of blockchain technology (Alazab, Alhyari, Awajan, &

Abdallah, 2021; Kamble, Gunasekaran, & Arha, 2018; Wong, Tan, Lee, Ooi, & Sohal,

2020). It has been determined that up to 40% variability in behavioural intention to use

technology could be explained by the combination of PEOU and PU (Lai, 2017; Tarhini,

Arachchilage, Masa'deh, & Abbasi, 2015). This supports the strong predictive power of

the constructs. Yet, TAM and its extensions also have drawbacks. First, it should be

remembered that TAM is predominantly a model focused primarily on an individual

user. TAM2 was intended to solve this issue by introducing additional job-related

variables such as job relevance and output quality (Venkatesh & Davis, 2000). Still,

these factors, even though within an organisational context, are applicable at an

individual level. In other words, TAM does not consider technology adoption as an

organisational goal to meet specific organisational needs. As such, it could be more

appropriately used to analyse how organisational members, as end-users of technology,

would accept and adopt it in the workplace. This is further confirmed by the absence

even in the latest TAM versions of some common environmental factors that could

influence organisational decision making in making technology investments (such as,

for example, regulations and the presence of financial resources). Finally, TAM is

mostly about the acceptance rather than the adoption of technology. Granted, the

endpoint of the framework is actual technology use. However, the influence of all

independent factors is on technology characteristics rather than on adoption or even

behavioural intention to adopt. For these reasons, TAM was not considered for this

study.

4.3.2.6 Diffusion of Innovations Theory

In many respects, Diffusion of Innovations Theory (DOI) (Rogers, Diffusion of

innovations, 1995) is considered an important adoption theory in many fields when it

comes to organisational level research (Lai, 2017; Tarhini, Arachchilage, Masa'deh, &

Abbasi, 2015). The theory emerged from general observations of common elements that

underlie diffusion research in several disciplines:

sociology, psychology, communications, economics and organisational life (Alalyan,

Jaafari, & Hussain, 2023; Rogers E. , 2003). Rogers formulated the DOI theory in his

original book Diffusion of Innovations in 1962. The theory, in its basic, aims to explain

the process of innovations spread and whether it has chances of being adopted (Fagan,

2001). Rogers (2003) defined diffusion of innovation as "the process in which an

innovation is communicated through certain channels over time among the members of a

social system” (p. 12). An innovation in its original formulation did not have to be

technological, although it turned out to fit well with the technology adoption in

organisational contexts (Khan & Qudrat-Ullah, 2020; Taherdoost, 2018).

Key Elements

Within DOI, four key elements of the theory can be identified: (1) the innovation itself,

(2) communication channels about innovation, (3) time it takes to spread, and (4) the

social system which provides the context.

The first element of the DOI is innovation. Rogers (2003) defined it as "an idea,

practice, or object that is perceived as new by an individual or another unit of adoption"

(p. 62). An important clarification made by Rogers is that innovation does not

necessarily mean something absolutely new, not known to anyone. Instead, innovation

knowledge differs among individuals and organisations. Within the DOI, innovation is

something that appears new to the target group. Therefore, with reference to this

particular study, blockchain is already an established technology in cryptocurrencies

whereas its novel applications in the higher education field may not be. It will be

considered an innovation then for those institutions that try to apply it for their own

purposes.

The second element of the DOI theory is communication channels. According to Rogers

(2003), "the diffusion process is the information exchange through which one individual

communicates a new idea to one or several others" (p. 18). This exchange is facilitated

by communication channels which could be as simple as personal interactions and as

global as programs on TV. The key point here is that diffusion is a social process which

is not possible without information about innovation being communicated. Rogers

argues that a successful communication channel for diffusion of innovation involves a

degree of heterophily, that is, differences between individuals in terms of certain

abilities (p. 19). This, for example, can help those who do not understand an innovation

fully to get new insights from those who do and can explain its benefits.

The third component of the DOI is time. Rogers (2003) argued that including time as a

key dimension to the diffusion of innovation allows it to be presented as a process. This

is also the main difference from the earlier theoretical frameworks of adoption. In DOI

theory, there are three major dimensions of the element of time. The first step is the

innovation decision process, which comprises five stages, beginning with the

individual's first understanding of the invention and ending with its adoption or

rejection. Next, innovativeness refers to a member's early or late adoption of innovation

within the same system. The third factor is the pace of innovation, which is often

evaluated by the number of members who adopt the innovation during a given time

period (Rogers, 2003). These elements are discussed in more detail below.

The final element of the DOI is the social system. This is "a set of interrelated units that

are engaged in joint problem-solving to accomplish a common goal" (Rogers, 2003,

p.37). A social system can be envisioned as an organised collective of units such as

individuals, groups, organisations, or societies. Since Rogers considered innovation

diffusion as taking place in social systems, he argued that the structure of such systems

would be influential in the process. A structure is defined within the DOI as “the

patterned arrangements of the units in a system” (Rogers, 2003, p. 24). These

arrangements, along with the shared problem-solving process, influence the degree of

innovativeness in a given social system. This assertion serves as a basis for

distinguishing adopters within the DOI.

Diffusion Process

According to DOI, technology adoption is a process, which goes through five stages

preceding the decision whether to adopt a technology (Figure 15). It starts with the

assertion that knowledge of the technology has to be present. Those who adopt the

technology should be aware of it and how it functions. The next stage is persuasion

where information about technology becomes more reliable as it comes from peers and

other technology users. At this stage, according to DOI, a decision borderline is drawn.

Next comes the decision stage where the choice of technology use is made. Importantly,

while a positive decision is formed at this stage, the technology can still be rejected at

any stage of the process. In this case, further stages are automatically eliminated. In case

of a positive decision regarding the technology, the implementation stage involves its

practical applications. At this stage, the technology is further assessed in terms of

usability, complexity and benefit. Finally, at the confirmation stage, adoption is

cemented.

Figure 15. Innovation-Decision Process Model (Rogers, 2003)

As seen from the model, the early stages of the diffusion process contain a number of

factors which, according to DOI, play an important role in moving towards the next

stages of adoption. DOI distinguishes among different decision-making units in the

knowledge stage: whether the decision will be made by individuals or organisations

(Rogers, 2003). Accordingly, the adopters will have different needs, skills, and possibly

experiences with technology. Further, in line with the behaviour-based theories (TRA,

TPB), DOI asserts that norms existing in the social system within which the adopter

operates, will also play a role.

The persuasion stage, on the other hand, contains the key innovation attributes that

influence the decision about its adoption. These five elements have become the

dominant technology factors in many adoption studies (Khan & Qudrat-Ullah, 2020;

Tarhini, Arachchilage, Masa'deh, & Abbasi, 2015). The first attribute is relative

advantage, which is “the degree to which an innovation is perceived as being better than

the idea it supersedes” (Rogers, 2003, p. 229). Relative advantage has different

expressions, and it can be seen in terms of cost, task completion or even social status

(Alalyan, Jaafari, & Hussain, 2023). The second attribute is compatibility, which is “the

degree to which an innovation is perceived as consistent with the existing values, past

experiences, and needs of potential adopters” (Rogers, 2003, p. 15). This innovation

aspect is considered important because different groups of adopters are likely to have

different values and beliefs (Alalyan, Jaafari, & Hussain, 2023). The third attribute is

complexity, which refers to “the degree to which an innovation is perceived as relatively

difficult to understand and use” (Rogers, 2003, p. 15). This is a negative attribute since

technologies which are difficult to learn and apply are likely to be adoptedmore slowly

(Alalyan, Jaafari, & Hussain, 2023). The fourth attribute is trialability, which is “the

degree to which an innovation may be experimented with on a limited basis” (Rogers,

2003, p. 16). Technologies that can be tried have a better chance of being adopted since

the potential adopters are likely to see hands-on positive benefits from them (Alalyan,

Jaafari, & Hussain, 2023). Finally, observability is “the degree to which the results of an

innovation are visible to others” (Rogers, 2003, p. 16). This attribute refers to the degree

of the innovation’s visibility and the visibility of the outcomes of its use (Alalyan,

Jaafari, & Hussain, 2023). Observable innovations that produce visible positive effects

are more likely to be adopted.

Innovativeness and Adopter Categories

According to Rogers (2003), diffusion of innovations is not a uniform process. This is

because individuals treat uncertainty related to innovations differently: some are ready

to try innovations earlier while others prefer to wait. Rogers (2003) referred this to

innovativeness: “the degree to which an individual or another unit of adoption is

relatively earlier in adopting innovative ideas than the other members of a system” (p.

22). Rogers (2003) proposed that in a typical society, there are several distinct groups

based on their degree of innovativeness (Figure 16).

The first group to adopt a new technology is known as innovators. This is a relatively

small proportion of adopters (2.5%) who try an innovation earlier than others. Rogers

(2003) compared innovators to system gatekeepers who express a high degree of

knowledge about innovations overall and who are keen to try them first. This group

differs from the others by having a strong capacity to understand the core ideas behind

new technologies and know how to deal with the uncertainties surrounding them. This

group initially paves the way for other groups by making innovations visible.

The second group to adopt the new technology is early adopters. They represent about

13.5% of the adopters who are eager to try innovations early. Rogers (2003) referred to

this group as change agents because it consists of influential individuals and opinion

leaders who usually model adoption for the later groups. Generally, early adopters are

very receptive to change and play a crucial role in the knowledge and persuasion stages

of the adoption process. In fact, they “play a central role at virtually every stage of the

innovation process, from initiation to implementation, particularly in deploying the

resources that carry innovation forward” (Light, 1998, p. 19). With this, early adopters

reduce much uncertainty about innovations and spearhead the adoption process for the

majority.

The third group of adopters is early majority. This is one of the largest groups (34% of

adopters), which consists of individuals who make adoption decisions after they observe

the innovation’s reputation. Unlike early adopters, who are opinion leaders, this group

consists primarily of opinion followers. Accordingly, Rogers (2003) referred to this

group as the deliberation group because they need the innovation to be visible and

require some time to absorb the information about its benefits. As soon as this become

available, the early majority group adopts the technology.

The fourth group is late majority. Similar to the early majority, this is a very large group

consisting of 34% of adopters. Late majority comprises individuals who are generally

sceptical about the innovation in question and the benefits it may produce (Rogers,

2003). Accordingly, this group adopts innovations when there is a substantial degree of

information about them and/or when the costs of adoption drop to a level which is

acceptable to them. Peer pressure and economic necessity may also be factors that

prompt this group to adopt (Rogers, 2003).

The final group of adopters is laggards. This is a fairly large group consisting of 16% of

innovation adopters. Rogers (2003) characterised this group as possessing the lowest

amount of resources and knowledge. They may not actually need the technology that

much until it becomes a standard, and this could happen even after newer innovations

take place.

Figure 16. Innovation Adopters’ Classification over Time (Rogers, 2003)

The classified types of adopters explain how and why a typical innovation becomes

distributed across time. There are a few early adopters who are ready to try innovative

ideas even at an increased cost and uncertainty. As the innovation spreads, it becomes

increasingly visible and the results of its use are observed. This prompts an increased

inflow of adopters, up until a time at which a critical mass of users is reached. After this,

adoption proceeds at an increased rate and becomes self-sustaining (Tarhini,

Arachchilage, Masa'deh, & Abbasi, 2015).

Theory Applications

DOI has been tested and applied in numerous studies and it remains one of the most

cited technology adoption theories (Khan & Qudrat-Ullah, 2020; Taherdoost, 2018). The

theory has been applied in numerous disciplines and industries to study the adoption of

emerging technologies. Some recent examples include the adoption of autonomous

vehicles in the transportation sector (Yuen, Cai, Qi, & Wang, 2021), the adoption of

exam monitoring systems by education institutions during the COVID-19 outbreak

(Raman, Vachharajani, & Nedungadi, 2021), the adoption of mobile wallets in the

financial sector (Shaw, Eschenbrennen, & Brand, 2022), and the adoption of the

internet-of-things in the healthcare sector (Yesmin, Carter, & Gladman, 2022) among

others. Prior research has also confirmed the validity of the five technology attributes’

effect in the technology adoption process (Khan & Qudrat-Ullah, 2020; Lai, 2017;

Tarhini, Arachchilage, Masa'deh, & Abbasi, 2015).

The aforementioned findings make DOI a suitable theoretical framework for this

research. As a strong organisational-level theory which is often used to study the

adoption of emergent technologies, DOI is an appropriate theoretical choice to study the

adoption of blockchain (an emergent technology) in higher education institutions

(organisational level). In fact, several studies have used DOI, either as a standalone

theory or in combination with other theories and frameworks, to study blockchain

adoption in various settings. The technology attributes within DOI, for example, were

investigated as antecedents of blockchain adoption in the freight industry (Orji, Kusi-

Sarpong, Huang, & Vazquez-Brust, 2020), sustainable supply chains (Kouhizadeh,

Saberi, & Sarkis, 2021), insurance (Kar & Navin, 2021) and smart learning

environments (Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020). Therefore, DOI

has already established certain ground in blockchain adoption studies across a number

of disciplines and industries. This provides another reason to use it as a theoretical

foundation for this research.

Limitations

Like any other theory of adoption, DOI is not without its own limitations. One of the

main criticisms of the theory is that it treats diffusion as a discrete package that takes

place in fixed, homogenous social environments (Lyytinen & Damsgaard, 2001).

Technological systems, especially complex systems, are usually interpreted and valued

differently across time and place (Khan & Qudrat-Ullah, 2020). Accordingly, local

traditions and norms, infrastructures, and socio-political and economic realities may

account for differences in technology adoption. This may be especially true for

organisations which are more susceptible to environmental arrangements than individual

users. Therefore, researchers proposed that analyses of technology diffusion and

adoption should dynamically draw from both environmental and institutional forces

(Robertson, Swan, & Newell, 1996; Wolfe, 1994). DOI pays less attention to these

forces than to the attributes of the technology itself.

Another line of DOI criticism is that, being primarily a communication theory, it treats

adoption choices as largely outcomes of the available information, the adopter’s

preferences and properties (Lyytinen & Damsgaard, 2001). Instead, the critics argue,

adoption choice parameters could be far more diverse. For example, in the case of

organisations, additional factors could be the chosen business strategy (does the

innovation support it?), management support and overall readiness for the innovation in

question (Clohessy & Acton, 2019; Khan & Qudrat-Ullah, 2020). A complex interplay

of these factors is, for example, described in the Business Ecosystem Model which

changes and evolves through the constant interaction of its economically linked

constituents (Moore J. F., 2006). These processes and dynamics then should not be

ignored in defining and scoping the innovation diffusion process (Lyytinen &

Damsgaard, 2001).

Finally, as a generalised theory, DOI treats the technology diffusion process as set

within fixed stages, relatively short time scales and it mostly ignores the effect of

previous decision-making processes (Lyytinen & Damsgaard, 2001). Accordingly, the

mechanism of innovation diffusion which drives adoption is treated as linear, fast and

more or less mandatory. In other words, the theory ignores possible feedback from the

system and considers both organisational and technology characteristics to be stable

over time (Khan & Qudrat-Ullah, 2020). However, this may not be the case, and the role

of process aspects, histories, and organisational evolution should be recognised (Ardis

& Marcolin, 2017).

To sum up, DOI as a standalone theory cannot account for the complexities of the

adoption process from the perspective of an organisation. While emphasising the role of

technology characteristics, it does not account for intra-organisational and

environmental factors sufficiently well. Further, it is often criticised for its linear nature

and the absence of feedback which could influence the adoption process. In order to

address these limitations, this study integrates DOI with the Technology-Organisation-

Environment (TOE) framework (Tornatzky & Fleischer, 1990). In doing so, the study

aims to provide a solid theoretical foundation for the research and develop a framework

which is both comprehensive and flexible.

4.3.2.7 Technology – Organisation- Environment Framework

Technology-Organisation-Environment (TOE) is a theoretical framework developed

specifically for studying the adoption and implementation of technology innovations by

organisations (Baker, 2012). The model was presented in Tornatzky and Klein’s (1990)

book Processes of Technological Innovations which offered a comprehensive review of

innovation adoption by firms. The book focused on the influence of context in which

firms operate on the process of technology adoption. More specifically, TOE proposes

that three such contexts play a role in a firm’s adoption of technologies: technology

context, organisational context and environmental context Figure 17. Each of these

contexts and their roles in technology adoption are described in detail below.

Figure 17: TOE framework (Tornatzky et al., 1990, p.154)

The Technology Context

The technology context within TOE includes technologies which are both in use and

available in the market which could be relevant to the organisation (Clohessy et al.,

2020). Two technology types are relevant for new technology adoption: the technologies

used by the organisation define the scope and limit of the changes that it could apply

whereas not used but available technologies define what is possible to achieve

(Tornatzky et al., 1990). Tornatzky and Klein (1990) distinguished three types of

innovations that define the adoption process based on the amount of risk they produce

for organisations. First, incremental change technologies either update the existing

technologies or introduce additional features to them. According to Baker (2012),

incremental change technologies pose the least risk to organisations because of their

gradual nature and lack of a groundbreaking effect on one or more aspects of

organisational operations. An example of such innovation is a shift towards LCD

monitors in PCs: the shift occurred over time and did not require immediate changes to

operations. Second, synthetic change technologies combine the existing technologies in

some new ways. According to Tornatzky and Klein (1990), these technologies pose a

moderate amount of risk for organisations because the effect of such combinations

usually takes time. An example of such innovation is online courseware, which

combined the existing product (education) with the internet to offer a new approach to

providing education services. Finally, discontinuous innovations represent radically new

ideas and innovations that change the way things are done (Baker, 2012). These changes

are of the highest risk for organisations because they represent fundamental shifts that

lead to new technology standards and the displacement of legacy technology systems.

Blockchain is an example of such innovation because it provides a completely new

approach to managing digital data.

The point of distinguishing technologies by type of risk, according to Tornatzky et al.

(1990), is in predicting the patterns of adoption. Specifically, for incremental change

technologies, a measured pace of adoption could be acceptable (Baker, 2012). The

synthetic change technologies require faster decision-making, although it could still be

possible to postpone adoption for better clarity with regard to its functions and benefits.

Finally, the discontinuous change technologies require fast and decisive decisions if

organisations want to remain competitive (Baker, 2012). This is especially true with

“competence-destroying” (Tushman & Nadler, 1986) technology solutions as they

render many existing systems obsolete. Therefore, according to TOE, organisations

must keep an eye on innovations and analyse the type of change expected from a

technology considered for adoption.

The TOE framework originally includes technology availability and characteristics as

the key factors in this dimension (Tornatzky et al., 1990). However, additional factors

have been successfully integrated and tested. The five technology attributes from DOI

are commonly considered within TOE (Clohessy & Acton, 2019; Lustenberger,

Malešević, & Spychiger, 2021). Other frequently explored technology context factors

are privacy and security, perceived benefits and technology maturity (Kouhizadeh,

Saberi, & Sarkis, 2021; Reddick, Cid, & Ganapati, 2019; Wamba, Queiroz, & Trinchera,

2020).

The Organisational Context

The organisational context involves a firm’s resources and characteristics that could

influence the adoption process (Baker, 2012). Tornatzky and Fleischer (1990) proposed

several mechanisms through which such influence is recognised. First, there are internal

individuals and groups that promote innovations, examples being innovation champions,

opinion leaders and technology gatekeepers. Second, innovations can be promoted by

cross-functional teams that have connections to organisational departments and partners.

Third, organisational structures that are more decentralised and team-oriented are more

supportive of innovations. Further, communication processes can either promote or

impede the innovation adoption process (Tornatzky & Fleischer, 1990). Within TOE, top

management plays a crucial role in this by either communicating support and linking

innovation to their organisation’s mission and vision or formally communicating why

the innovation does not fit with the organisational strategy.

The impact of the two remaining factors within the original TOE, slack and

organisational size, is much less certain in relation to technology adoption. While

organisational slack has been cited as supporting innovations in some classical

organisational works (March & Simon, 1958) and even within DOI (Rogers, 1995), the

presence of this factor is not indicative of adoption. Tornatzky and Fleischer (1990)

acknowledged that it is “neither necessary nor sufficient for innovation to occur” (p.

161). Similarly, while larger organisational size has been traditionally associated as

better suited to adopt innovations, later studies painted a rather mixed picture on this

(Baker, 2012; Clohessy & Acton, 2019). It was, therefore, argued that size serves as a

very crude approximation of more specific resources required to stimulate adoption

(Baker, 2012). In view of the uncertainties surrounding these two factors, later research

introduced additional organisational variables such as, for example, organisational

readiness and innovativeness (Guo & Liang, 2016; Morabito, 2017; Pilkington, 2016;

Woodside, Augustine, & Giberson, 2017).

The Environmental Context

The environmental context represents the ecosystem within which the organisation

operates: the environment with which it interacts but has little to no power to influence.

Lippert and Govindarajulu (2006) described this context as a “setting in which the firm

conducts business, and is influenced by the industry itself, its competitors, the firm's

ability to access resources supplied by others, and interactions with the government” (p.

149). According to Tornatzky and Fleischer (1990), the environmental mechanisms have

a number of mechanisms that influence innovation adoption. It is proposed, for

example, that more intense competition stimulates novel technology adoption to gain a

competitive advantage. Likewise, dominant partner firms may stimulate organisations to

adopt innovations to maintain business relationships and enhance cross-organisational

operations. Further, according to TOE, adoptions occur faster in environments with a

good supply of technologies and appropriate technology infrastructures (Baker, 2012).

Finally, all firms, regardless of their adoption intent, exist and operate within legal

environments created and maintained by governments. Existing regulations may either

stifle innovations by imposing a high level of costs or restrictions on the technology in

question or mandating adoption by making some obligatory (one example would be

antipollution systems).

Similar to the other two dimensions within the original TOE, the environmental context

has also been modified in many studies to add or remove certain variables. Government

support, for example, emerged as one of the commonly included variables in the

environment context (Crosby, Nachiappan Pattanayak, Verma, & Kalyanaraman, 2016;

Guo & Liang, 2016; Tapscott & Tapscott, 2016). Clohessy et al. (2020) identified other

variables that were tested to a lesser extent: business use cases, critical user mass,

democratisation and political stability.

To sum up, the three contexts introduced within TOE represent influential dimensions

for adoption decisions at the organisational level. As shown, the dimensions include

factors that could either stimulate or impede technology adoption by firms. Importantly,

TOE does not constrain these dimensions in terms of the factors included. This has

given researchers a good opportunity to explore additional contextual factors of

adoption within each type of environment. Such flexibility is beneficial for the current

study where new contextual factors are likely to emerge.

Framework Applications

Like DOI, TOE is one of the most commonly used frameworks in adoption research at

the organisational level (Clohessy, Treiblmaier, Acton, & Rogers, 2020; Lustenberger,

Malešević, & Spychiger, 2021). The framework is valued for its comprehensiveness and

ease of expansion/modification (Clohessy & Acton, 2019; Clohessy, Treiblmaier, Acton,

& Rogers, 2020). Zhu and Kraemer (2005) noted that TOE offers researchers much

freedom in adding or eliminating factors that they would consider relevant, which also

means that there is little need to modify the underlying theory. Baker (2012) also wrote

that TOE blends well with similar theories like DOI which makes it possible to integrate

them instead of offering competing views on the adoption process. Finally, TOE is

considered well fit for studying the adoption of new and emerging technologies

(Clohessy, Treiblmaier, Acton, & Rogers, 2020; Oliveira & Martins, 2011).

Recently, there has been an explosive growth of papers using TOE-related factors to

study blockchain adoption in various industries. A review by Clohessy et al. (2019)

identified 16 studies of such kind, while a year later, Clohessy et al. (2020) identified

31, thereby indicating a growth of almost 100%. Further, the TOE framework itself has

become rather popular in studying blockchain adoption (Chittipaka, Kumar, &

Sivarajah, 2022; Kamarulzaman, et al., 2021; Kulkarni & Patil, 2020; Schmitt,

Mladenow, Strauss, & Schaffhauser-Linzatti, 2019). Still, there are very few works that

applied TOE to study blockchain adoption in higher education (Barnes & Xiao, 2019;

Clohessy & Acton, 2019; Janssen, Weerakkody, Ismagilova, Sivarajah, & Irani, 2020).

Therefore, there is a need for more testing of the framework and checking its

applicability in different higher education contexts.

4.3.2.8 Enhanced DOI-TOE Framework Choice

This study uses an enhanced DOI-TOE framework as the theoretical basis for research.

The choice of such a framework is justified for a number of reasons. First, such a

framework clarifies the adoption of innovative technology in higher education

institutions (HEI) at the organisational level. As the review of the various adoption

models and theories shows, TRA, TPB TAM, TTF, SCT, TAM and its variations focus

primarily on adoption at an individual level. Indeed, studies of blockchain adoption

using these theories explored the importance of individual user factors such as attitudes,

comfort levels, performance expectancy, and personal perceptions of blockchain

(Alaklabi & Kang, 2021; Kamble, Gunasekaran, & Arha, 2018; Queiroz & Wamba,

2019; Wong, Tan, Lee, Ooi, & Sohal, 2020). While such factors are important in

understanding how blockchain adoption diffuses in societies, they explain little with

regard to the elements within and outside organisations that aim to introduce blockchain

for their business purposes.

In contrast, both DOI and TOE are organisational-level frameworks, as previously

discussed. Both theories represent solid research frameworks applied for the adoption of

various innovations across numerous disciplines and industries. Importantly, DOI and

Research subquestions 1, 2 and 3 explore the technology, organisational and

environmental factors promoting blockchain adoption in HEIs. In fact, different

combinations of factors within DOI and TOE have already been explored in the

blockchain adoption literature. Table 17 lists the empirical studies on blockchain

adoption in the educational sector that explored the influence of technological,

organisational and environmental factors identified within the DOI and TOE

frameworks.

Table 17: Studies of Blockchain Adoption with Factors from DOI and TOE

Study DOI Factors

Explored

Organisational

Factors Explored

Environmental

Factors Explored

Additional

Factors

Explored

Barnes & X

(2019)

iao Relative advantage

Compatibility

Complexity

Top management

support

Readiness Size

Centralisation

Competition

Government

support

Partner support

Technology vendor

support

Customer support

Industry sector

Chen et

(2018)

al. Complexity Top manage

support

Readiness

ment Government

support

Industry pressure

Market dynamics

Perceived

benefits Energy

consumption

Choi et

(2020)

al. Complexity

Compatibility

Knowledge

expertise

Collaboration

Technology

awareness

and Government

support

Regulations

Infrastructure

Privacy

security

Costs

and

Crosby et al.

(2016)

Relative advantage

Complexity

Top management

support Readiness

Size

Customer relationships

Government

support

Competition

Regulations

Partner’s pressure

Privacy

security

Perceived

benefits

and

Duan et

al. (2020)

Complexity

Compatibility

Top management

support Innovativeness

Competition Perceived

benefits

Transparency

Guo & Liang

(2016)

Relative advantage

Compatibility

Complexity

Top

management support

Readiness

Knowledge

Competition

Partnerships

Regulations

Business use cases

Security and

privacy Cost

Business concerns

Hartley et al.

(2021)

Relative advantage

Compatibility

Complexity

Top

management support

Size

Regulations

Partners’ support

Industry group

membership

Institutional

factors

System update

Consulting

Iansity &

Lakhani

(2017)

Relative advantage

Complexity

Trialability

Compatibility

Top

management support

Readiness

Size

Competition

Partnerships

Regulations

Business use cases

Savings

Accessibility

Kouhizadeh et

al. (2021)

Complexity

Compatibility

Top management

support Readiness

Competition

Regulations

Market standards

Security

Perceived benefits

Network

incentives

structure

Lustenberger et

al. (2021)

Relative advantage

Compatibility

Complexity

Trialability

Observability

Top management

support Readiness

Size

Age

Competition

External pressure

Regulations

Collaboration

Ecosystem scope

Blockchain

knowledge

Malik et al.

(2021)

Complexity

Compatibility

Top management

support Innovativeness

Learning capability

Competition

Government

support

Partners’ readiness

Standardisation

Perceived risks

Perceived

benefits

Morabito (2017) Complexity

Compatibility

Top management

support Readiness

Size

Regulations

Government

support

Perceived

benefits Costs

Innovativeness Partner pressure

Business use cases

Pilkington

(2016)

Complexity

Compatibility

Top management

support Readiness

Size

Innovativeness

Network incentive

structure

Competition Perceived

benefits

Blockchain type

Toufaily et al.

(2021)

Complexity Readiness

Business model

alignment

Regulations

Network effects

Ecosystem

readiness

Privacy and

security Costs

Wamba et al.

(2020)

Complexity Top management

support Innovativeness

Competition Perceived

benefits

Transparency

Wang et al.

(2016)

Complexity

Compatibility

Top management

support Readiness

Size

Responsiveness

Regulations Industry

pressure

Market dynamics

Security,

uncertainty,

maturity of

technology

Woodside

et al.

(2017)

Complexity

Compatibility

Readiness

Innovativeness

Regulations

Market dynamics

Security

Perceived

benefits Costs

Zheng et al.

(2018)

Complexity Top

management support

Readiness

Size

Innovativeness

Regulations

Market standards

Business use cases

Security

Perceived

benefits

Based on the review of the theoretical and empirical literature, the following factors were

included in the model:

- Technology Factors: relative advantage, observability, compatibility, complexity

and trialability;

- Organisational Factors: top management support,

organisational size, organisational readiness;

- Environmental Factors: regulations, government support, peer pressure.

4.4.2 Solution to Research Subquestion 4

While DOI and TOE offer a rather comprehensive view on adoption at the

organisational level, neither of the theories considers an important aspect of higher

education, which is quality. Quality refers to “the excellence, standards, perfection,

conformance to requirements, fitness for purposes and value for money of the

educational technology services level and higher education institutions outcomes”

(Harvey & Knight, 1996, p. 13). Higher institutions can adopt ICT when acquiring a

relative advantage. Universities and colleges in developing countries, such as Saudi

Arabia, can improve their educational quality by adopting the latest technology, such as

blockchain technology (Duan, Zhang, Gong, Brown, & Li, 2020; Farah, Vozniuk,

Rodriguez-Triana, & Gillet, 2018; Xu, et al., 2017). Adopting blockchain technology,

increasing data security, and privacy-enhancing trust, reducing costs and improving

efficiency and immutability, adopting a ubiquitous global database, and incorporating

formative evaluation are highly beneficial to improving education mechanisms

(Alammary, Alhazmi, Almasri, & Gillani, 2019; Kolvenbach, Ruland, Grather, & Prinz,

2018).

Higher education quality can be improved by incorporating a variety of approaches,

such as quality assurance and quality enhancement, as identified by Lomas (2004).

Quality assurance pertains to measures taken to avoid producing subpar products, while

quality enhancement endeavors to enhance the level of students' education. In the realm

of higher education, the impact of quality outcomes is extremely important as the

education of every nation has a bearing on its economy and the global economy at large.

In most cases, there is a direct relationship between higher education outcomes and

social and economic development (Harman & Meek, 2000).

Table 18 presents five factors that define quality as drawn from the relevant research on

quality in higher education (Al-Ramahi & Odeh, 2020). The first factor is excellence in

ICT services which refers to exceeding the minimal required service standards in higher

education. The second factor is ICT perfection which refers to minimising process

limitations such as, for example, delayed response time, service downtime, and others.

The third factor is value for money, which is defined as the level of return on technology

investment. In the context of higher education institutions, this usually refers to

improvements in service quality and operational efficiencies. The fourth factor is fitness

for purpose, which essentially means that ICT has to fit the prescribed goals and

objectives in education quality improvements. The final factor is higher education

institution outcomes, which means improvements in university outcomes for the

institution itself and its stakeholders.

Table 18. Indicators of Quality in Higher Education

Quality Indicator Definition Key Concepts

Excellence High achievement standard

(Harvey & Stensaker, 2008)

Exclusivity, exceeding

expectations, achieving more than

required

Perfection Minimisation of process

limitations (Harvey, 2007)

Zero defects, doing right from the

first time, culture of quality.

Value for Money Returns on technology investment Service quality improvement,

(Harvey & Knight, 1996) process efficiencies,

accountability

Fitness for Purpose

Ability to accomplish goals and

objectives (Harvey & Green,

1993)

Meeting specifications, fitting

organisational mission, goals,

objectives

Higher education

institution outcomes

Ensuring positive technology

effects for the institutions and their

stakeholders (Harvey, 2006)

Enhancing processes, empowering

students, spurring innovations,

meeting labour market demands

Including the contribution of blockchain to quality improvements in this research is

justified in several ways. Higher education institutions are keen to improve the quality

of their services since this is usually a point of differentiation and a competitive

advantage (Ham & Hayduk, 2003; Tsinidou, Gerogiannis, & Fitsilis, 2010; Waller,

Lemoine, Mense, & Richardson, 2019). Innovative technology, in turn, is often seen as a

way to enhance service quality in higher education (Danjum & Rasli, 2012; Pavel,

Fruth, & Neacsu, 2015). Empirical research, for example, shows that the adoption of

relatively recent technologies like cloud computing, mobile learning, and virtual reality

have substantially advanced higher education service outcomes (Crompton & Burke,

2018; Qasem, Abdullah, Jusoh, Atan, & Asadi, 2019; Radianti, Majchrzak, Fromm, &

Wohlgenannt, 2020). There is also growing evidence that blockchain adoption has

brought service improvements in healthcare, supply chains, and other sectors (Loizou,

Karastoyanova, & Schizas, 2019; Tandon, Dhir, Najmul Islam, & Mäntymäki, 2020;

Tijan, Aksentievic, Ivanic, & Jardas, 2019). Finally, this study suggests one of the

reasons to add quality as a factor is the high unemployment level in Saudi Arabia, which

can be directly linked with the education quality of its universities and colleges.

Therefore, the following factors are included in the quality dimension: reduction in

graduates’ unemployment, education service improvements and HEI administration

improvements.

4.4.3 Solution to Research Subquestion 5

Neither DOI nor TOE distinguish barriers to adoption as a separate category of factors.

However, the successful adoption of blockchain in Saudi HEIs will certainly require

identifying such barriers so that they can be successfully addressed and overcome.

Granted, certain barriers to adoption are presented in both DOI (technology complexity)

and TOE (regulations). However, they remain predominantly oriented towards enablers

and consider barriers to adoption among them, not separately. Choi et al. (2020) argued

that this is a serious omission, especially when it comes to applying these frameworks to

study uncertain technologies like blockchain.

Given the fact that, unlike some other industries, higher education is relatively slow in

the adoption of blockchain (Alammary, Alhazmi, Almasri, & Gillani, 2019; Ullah,

AlRahmi, Alzahrani, Alfarraj, & Alblehai, 2020), it is logical to assume that there are

inherent hurdles that prevent the diffusion of this technology. Therefore, for

practitioners, introducing a separate category of factors that impede blockchain adoption

in Saudi HEI could be of great importance. Moreover, barriers to blockchain adoption in

education have been well explored in literature. As shown in Section 2.6, researchers

have identified at least 8 technological, 7 organisational and 5 environmental barriers to

blockchain adoption in higher education. Moreover, additional barriers could be present

in the context of Saudi HEIs which need to be further explored. For these reasons, this

study integrated a separate barriers dimension into the DOI-TOE framework.

Based on the literature review, the following factors are included in the barriers

dimension: lack of knowledge, privacy and security concerns, risk avoidance, lack of

infrastructure, lack of finance, lack of specialists, lack of technology visibility.

4.5 Methodology for the Solution Test and Evaluation

After the model is finalised, it has to be tested. Specifically, the factors included in the

model have to be validated and their relationships to blockchain adoption have to be

confirmed. A set of specific procedures used to test the relationships within a framework

is known as a research methodology (Creswell & Creswell, 2017; Wiles, Bengry-

Howell, Crow, & Nind, 2013). The choice of methodology guides the way for specific

methods and approaches to data collection and analysis, which are necessary to validate

the model (Bryman, 2016). Saunders et al. (2019) visualised research methodology

within a “research onion” with a series of layers leading from broader design choices

such as a paradigmatic or philosophical basis to more specific methods for data

collection and analysis (Figure 19).

Figure 19: The 'Research Onion' (Saunders et al., 2019, p. 108)

4.5.1 Methodological Choice

Two major methodological strands are recognised in the literature: qualitative and

quantitative. Qualitative research is often rooted in interpretivist philosophy inductive

reasoning (Creswell J. W., 2014). Researchers who choose qualitative methodologies

often seek to submerge themselves in specific contexts and gain a deep understanding of

the study phenomena (Neuman, 2006; Olfazoglu, 2017). Data collection and analyses

are oriented towards the development of meanings and perspectives with a strong focus

on the study participants. This allows what is being studied to be described thoroughly

and comprehensively. There are a number of methodological choices for this, the major

six being phenomenology, ethnography, grounded theory, action research, case study

and narrative research (Creswell and Creswell, 2017). Despite the richness and quality

of the collected data and the flexibility of its interpretation, qualitative methods are

criticised for their subjectivity, researcher bias, and difficulty in generalising the results

(Olfazoglu, 2017; Taylor & Trumbull, 2005). Quantitative research usually arises from

the positivist worldview and it uses deductive reasoning as the main approach to theory

building (Bryman, 2016; Creswell and Creswell, 2017). Quantitative approaches apply

numerical methods to study and explain particular phenomena. The emphasis is on the

objectivity of the data collection and analysis techniques, high levels of data reliability

and validity and large sampling sizes which enable the results to be generalised. Some

common quantitative methods in research include experiments, quasi-experiments,

surveys and correlational studies (Creswell and Creswell, 2017; Taylor and Trumbull,

2005). Quantitative methodologies are praised for their high level of precision, relative

ease of replication and straightforwardness in the explanation of the findings. The main

criticisms of quantitative methodologies relate to an excessive focus on numbers at the

expense of broader themes and underlying topics, a reliance on standardised procedures

that may not be fitting in specific contexts and the static presentation of the results

(Bryman, 2016; Tashakkori, Johnson, & Teddlie, 2020).

This study seeks to determine the factors that influence blockchain adoption in Saudi

HEIs. Currently, blockchain adoption in Saudi Arabia’s higher education sector remains

mostly an unexplored area. Saudi researchers have so far concentrated on the reviews of

blockchain applications in higher education (e.g., Alammary et al. 2019, Alam and

Benaida, 2020, Malibari, 2020) whereas empirical studies are virtually non-existent.

However, blockchain adoption studies in education institutions in other countries are

growing rapidly. Further, DOI and TOE are commonly established theories that are used

to investigate blockchain adoption factors in the education sector (Barnes and Xiao,

2019; Clohessy et al., 2020; Janssen et al., 2020; Toufaily et al., 2021; Ullah et al.,

2020). Therefore, an effective approach to meet the study purpose is to combine

qualitative and quantitative methodologies to capitalise on the strengths of each.

Qualitative research can help understand the topic of blockchain adoption in education

applied specifically to the Saudi HEIs context and quantitative research can offer a

means by which to test the established theory and relationships in this context.

Therefore, a mixed methods methodology was chosen for this study.

4.5.1.1 Rationale for Use

Mixed methods research is defined as research that “combines elements of qualitative

and quantitative research approaches … for the broad purposes of breadth and depth of

understanding and corroboration” (Johnson, Onwuegbuzie, & Turner, 2007, p. 123).

This is different from multi-method qualitative or quantitative studies where more than

one method of the same methodology type is applied (Saunders et al., 2019). In this

way, mixed methods research integrates the perspectives attained in both qualitative and

quantitative methodologies while reducing their weaknesses. In fact, overcoming the

limitations of each methodology was the reason for introducing mixed methods research

in the first place (Campbell & Fiske, 1959; Webb, Campbell, Schwartz, & Sechrest,

1966).

The mixed methods methodology is being increasingly recognised today as a superior

approach to research in contrast to pure qualitative or pure quantitative methods

(Tashakkori et al., 2020). Creswell and Creswell (2017) argued that the mixed methods

methodology enhances a researcher’s understanding of the phenomena being

investigated. It is also argued that a mixed methods approach can enhance the validity of

the study results by combining and comparing the findings from quantitative and

qualitative analyses (Tashakkori et al., 2020). Collins et al. (2007) proposed that by

using a mixed methods methodology: 1) the sample size can be balanced with in-depth

perspectives from the participants; 2) the validity of the data collection instruments can

be improved; 3) high levels of data integrity can be attained; and 4) the study findings

can be enhanced. Finally, Bryman (2016) argued that a mixed methods methodology

allows both processual and static information to be effectively investigated and for

different aspects of the investigated problems to be addressed.

By choosing a mixed methods methodology, this study aims to come to a detailed,

comprehensive understanding of blockchain adoption in Saudi HEIs which will allow

for robust theory building and testing. Mixed method studies of blockchain adoption are

not as widespread as pure quantitative or qualitative studies, although there is an

indication that this type of methodology is gaining attention with several recent

publications (Clohessy & Acton, 2019; Delghani, Kennedy, Mashatan, Rese, &

Karavidas, 2022; Gökalp, Gökalp, & Gökalp, 2020; Werner, Basalla, Schneider, Hays,

& Brocke, 2021). Researchers conducting these studies criticised the lack of an

integrative approach to understanding blockchain adoption in specific contexts. Further,

the aforementioned papers took TOE as a theoretical basis but uncovered the context-

specific factors of blockchain adoption and tested them successfully as a result.

Therefore, there is good supportive ground for applying the mixed methods

methodology in the analysis of blockchain adoption, as intended in this study.

4.5.1.2 Choice of Mixed Methods Design

There are several ways to construct mixed methods research considering that it

combines quantitative and qualitative methods. Tashakkori et al. (2020) identified three

major types of mixed methods research: concurrent, sequential and conversion.

Concurrent mixed methods studies apply qualitative and quantitative methods

simultaneously and compare the results. Sequential mixed methods studies begin with

one methodology and then use the other one to clarify or enhance the results afterwards.

Two methodologies are possible here, where either the qualitative or quantitative

method takes precedence. Finally, the conversion design mixes both methodologies at

every research stage as the data are transformed for qualitative and quantitative

analyses. As such, four major types of the mixed methodologies are recognised, as

discussed below.

In the convergent parallel mixed methods design, researchers collect and analyse data

with quantitative and qualitative methods independently, at the same time (Creswell and

Creswell, 2017). Both types of analysis are given equal consideration, and the results are

compared to enhance the validity of the conclusions. This type of design is often used to

acquire an in-depth understanding of a phenomenon being investigated (Tashakkori et

al., 2020). In a sequential exploratory design, a qualitative methodology is prioritised:

after the qualitative data collection and analysis, quantitative methods are used to clarify

the findings or test/confirm the developed relationships (Creswell and Creswell, 2017).

This methodology is suited for developing new or refining existing theories and

perspectives and then testing them. In a sequential explanatory design, a quantitative

methodology is prioritised: the results of the experiments or statistical tests are used as a

foundation for knowledge building with the qualitative methods serving to clarify

certain results (Tashakkori et al., 2020). Finally, the embedded mixed methodology

assumes data collection is undertaken using both qualitative and quantitative methods,

although one method serves as the dominant, larger design whereas the other plays a

complimentary, supportive role (Creswell J. W., 2014). This methodology is applied to

enhance the findings acquired via the dominant methodology. The four mixed

methodology approaches are summarised in Table 19: Mixed Methodology Designs.

Table 19: Mixed Methodology Designs

Methodology Sequence Description Application

Convergent

parallel

Quant and qual done

simultaneously

The methodologies are

given an equal weight in

investigating and

explaining a

phenomenon

(triangulation).

Attain a thorough

understanding of a

research question and

enhance the validity of

the findings.

Exploratory

sequential

Qual first,

afterwards

quant Qualitative takes

precedence, quantitative

is used to test and

confirm the findings.

Develop/revise theories or

models and test them.

Explanatory

sequential

Quant first,

afterwards

qual Quantitative takes

precedence, qualitative is

used for further

explanations.

Gain an understanding

of the unexpected results

of experiments or

surveys.

Embedded Primary and

secondary

methodology.

Sequence depends on

the research needs.

One methodology is

dominant, the other

serves to support/clarify

research findings as it

progresses.

Improve the research

design, clarify the

findings at each stage of

the analysis.

(Adapted from: Creswell, 2017; Tashakkori et al., 2020)

This study applies the exploratory sequential design to investigate the research

questions. The main goal of the study is to identify the essential factors influencing

blockchain adoption in a specific context of Saudi HEIs. The context-specific approach

assumes that, while a general adoption theory may be applicable, some new factors may

emerge and literature-proposed factors may not be significant. To understand which

factors could be important, qualitative research was undertaken first where the expert

study participants offered insights about the applicability of the blockchain adoption

factors identified in the relevant literature and suggested new ones. After the initial

model was refined, a largescale quantitative study was undertaken to test the

relationships within the revised framework. The exploratory sequential study design is

visualised in Figure 20.

study participants closed and open-ended questions (Brannen, 2017; Fowler,

2013). This is an important feature given the mixed methodology choice in this study.

Third, surveys are well-established in the studies of human behaviour, psychology and

drivers (Singleton & Straits, 2009). This makes them well suited to examine the factors

driving blockchain adoption in education. Finally, surveys are notable for their reach, as

they allow both small and large populations to be covered with relative ease and provide

a possibility to generalise the results (Creswell & Creswell, 2017). As such, the choice of

the survey strategy enables research to be conducted both within a smaller group of

experts and a larger population of IT professionals and HEI administrators. It also helped

to overcome the time and budget constraints of this research.

4.5.3 Time Horizon

The time horizon layer defines the timeframe within which the study is conducted

(Saunders et al., 2019). The two alternative time horizons are longitudinal and

crosssectional. A longitudinal time horizon, as the name suggests, involves data

collection over a long period of time with the subsequent analysis and comparison of the

findings (Bryman, 2016; Creswell & Creswell, 2017). In contrast, a cross-sectional time

horizon assumes data collection over a short period of time to capture the state of

matters at a particular moment (Bryman, 2016; Creswell & Creswell, 2017). It attempts

to match the study population closely and, therefore, involves groups of different

individuals. While longitudinal research may offer more robust results and insights into

how variables are changing over time, they often demand serious investments in time

and resources, both human and financial, to meet the high rigour of the requirements

(Bryman, 2016). Further, longitudinal studies require a high degree of control as

participants tend to withdraw or change their behaviours. Finally, such studies are better

suited for investigating the influence of a few significant factors over time rather than

multiple factors acting simultaneously, as is the case with this study. For all the

aforementioned reasons, the cross-sectional time horizon is chosen for this research.

4.5.4 Methods for Data Collection and Analysis

As previously discussed, this study follows the exploratory sequential methodology. As

such, the data collection and analysis are conducted in two phases. First, qualitative

research is used to clarify the research frameworks and the relationships within it. After

this, quantitative research is used to test the updated framework and relationships. The

methods of data collection and analyses for each are described in detail below.

4.6 Research Phase I: Model Presentation

Following an extensive literature search and analysis, this study identified a series of

factors that the previously conducted research found significant in the process of

blockchain adoption in HEIs. However, to make the proposed framework more fitting to

the context of Saudi HEIs, qualitative research was undertaken. The purpose of

qualitative data collection and analysis is to refine the model by introducing the new

context-specific factors and eliminating ones that are considered insignificant.

4.6.1 Data Collection

Qualitative data collection was undertaken through a series of semi-structured

interviews. This approach is applied to obtain respondents’ views, opinions, thoughts

and perspectives on a specific topic (Creswell & Creswell, 2017). Such interviews

involve responses to open-ended questions that are organised around certain topics.

Hence, they are more flexible than structured interviews, making them suitable for

acquiring new perspectives on the issues while remaining within the subject narrative

without deviating too much from the important topics. To fulfil these objectives, semi-

structured interviews follow a specific protocol which contains subject topics/themes

and associated questions (Brinkmann & Kvale, 2015). The questions, in turn, serve as

guidance and may be asked in a different manner to different individuals.

The choice of semi-structured interviews for the qualitative data collection in this

research was based on several factors. The interviews were conducted based on a

previously developed framework and thus, can be easily organised by themes. It is

important to remain within these themes to follow the theoretical lines of the research.

On the other hand, a certain amount of flexibility is necessary to allow the respondents

to express ideas about new factors relating to blockchain adoption. As such, semi-

structured interviews are an ideal choice by using themes and allowing free expression

of opinions and insights within them. Most of the weaknesses of semi-structured

interviews, such as issues with generalising the findings and issues with subjectivity are

resolved by testing the findings with the statistical methods in the second phase of the

research.

Unlike quantitative types of research, there are no universally established norms for

sampling sizes in qualitative studies. The qualitative research literature (Guest, Namey,

& Chen, 2020; Fusch & Ness, 2015) and the literature specialising in qualitative

interviews (Brinkmann & Kvale, 2015; Weller, et al., 2018) generally recommend a

saturation approach. Data saturation is defined as a situation when additional data

collection does not yield meaningful new insights required for a robust understanding of

the studied phenomenon (Guest et al., 2020). Following these guidelines, this study

assesses the data saturation point from the perspective of developing new themes and

insights about blockchain adoption in Saudi HEIs. If no such insights are developed

over the course of a given interview, the data saturation point is reached.

A minimum number of interviewees nevertheless is determined to ensure that the

sample size is balanced in terms of gender, position and the type of HEI (private/public,

small/medium/large). To do this, a convenience sampling approach was used. Whereas

convenience sampling is not recommended in quantitative research, qualitative studies

often use it for the purpose of balancing a small sample to include possible

representatives of all study groups (Brinkmann & Kvale, 2015). The participants were

recruited through the personal network connections of the researcher. The main

requirement for inclusion in the study sample was that the participant should hold a

decision-making position in a higher education institution. The final sample included 10

individuals in information technology and management jobs in Saudi colleges and

universities.

The qualitative data collection took place in March 2022 using online communication

tools (Zoom, Skype). Prior to the interviews, all participants were given information

about the study purpose, the goals of the interview and their rights as the study

participants. In line with the literature recommendations, a set of themes was developed

to guide the interviews: overall views of blockchain potential for education and the

respondents’ HEIs, factors supporting blockchain adoption in education (grouped within

technological, organisational, environmental, human, resources and quality dimensions)

and barriers to adoption. The interview protocol is presented in Appendix A. The

interviews were conducted in Arabic and then translated into English, and the interview

times ranged from 30 to 50 minutes. Upon completion of the interviews, the participants

were sent their respective transcriptions so that they could make changes if necessary to

better express their thoughts and ideas. The revised transcripts were then used for the

data analysis.

4.6.2 Data Analysis

The collected qualitative data were mamually analysed using content analysis, which is

a common technique for analysing transcript data (Elo & Kyngäs, 2008; Guest,

MacQueen, & Namey, 2012). A directed content analysis technique is used in this study

because it is considered the best approach to build upon the existing theories and

frameworks (Assarroudi, Heshmati, Armat, Ebadi, & Vaismoradi, 2018; Hsieh &

Shannon, 2005). The goal of directed content analysis is “to validate or extend

conceptually a theoretical framework or theory” (Hsieh & Shannon, 2005, p. 1281),

which is the purpose of this qualitative research phase, namely seeking to refine the

framework for the existing adoption factors with a strong theoretical background by

adding and removing the factors emerging during the interviews.

The data analysis is based on the approach prescribed by Miles and Huberman (2019),

consisting of three stages. The first step in the qualitative data analysis involves data

reduction and organisation, which is followed by pattern coding and, finally, data

display and interpretation. The choice of the directed content analysis approach enables

data analysis to follow a more structured process than conventional approaches (Hickey

& Kipping, 1996). In the data reduction stage, the transcribed data are simplified and

coded. In the conventional content analysis approach, the codes are developed anew;

however, a directed content study usually uses the applied theories and models to

develop initial coding categories (Potter & Levine-Donnerstein, 1999). Accordingly, this

study has predetermined codes arising from the variables in the proposed conceptual

framework (Section 4.4 of this thesis). Whenever data cannot be readily assigned to any

predetermined category, they are assigned to a new category with the potential to be

developed into a novel blockchain adoption factor.

Following the data reduction and coding, patterns in the data were investigated and

marked. Specifically, the frequency of mentions for each coded factor were noted. The

factors that were mentioned frequently and the factors that were mentioned rarely were

investigated separately. The frequently discussed factors were analysed with regard to

potential influence on blockchain adoption, in terms of both direction (positive or

negative) and strength (weak, moderate, strong or unclear). The factors that were

discussed rarely were considered as candidates for elimination or integration with the

other factors. This is known as deductive category application (Mayring, 2000). Similar

patterns identified across the transcripts were used to develop the themes which would

help decide regarding a factor (retain in the model, eliminate, move/integrate with

others) and offer explanations for that decision.

The outcome of the qualitative analysis is the revised framework for blockchain adoption

in Saudi HEIs. The process of qualitative data analysis is summarised in Figure 21.

Figure 21: Qualitative Data Analysis Approach

4.7 Research Phase II: Model Evaluation

Quantitative methods allow researchers to collect and analyse large amounts of data

from many subjects, thereby increasing the chance of generalising the results (Creswell

J. W., 2014). Further, quantitative methods offer a higher degree of accuracy and

objectivity which ensure the reliability and validity of the findings. Through the use of

numeric analysis and statistical methods, quantitative studies make it possible to

compare these findings with the existing studies on the subject as well as facilitate

analyses across time and categories.

This study uses an online survey as the primary method of data collection. Surveys are

one of the most common approaches to data collection in quantitative research (Fowler,

2013; Tashakkori, Johnson, & Teddlie, 2020). Questionnaires are usually used as the

data collection tool (Creswell & Creswell, 2017). Online surveys have become common

in view of the development of digital and online technologies as well as the penetration

of the internet into wider society groups. The following advantages of online surveys

Interview Transcripts

Data Reduction

Data Coding

Verified Pre-Determined Categories

New Categories

Frequently Mentioned

Rarely Mentioned

Eliminated from the

Model

Moved/Integrated with

Other Factors

Theme Development

Predicted Factor

Strength

Predicted Relationship

Direction

Data Display

make them the preferred method for data collection in this study (based on Creswell,

2017; Saunders et al., 2019):

1. Structure and organisation: surveys can be easily replicated and the results can

be easily compared;

2. Large number of participants: it is possible to collect data from many

participants to ensure a large sample size involving individuals from various

backgrounds;

3. Reducing bias: online surveys detach the participants from the researcher thereby

minimising personal biases in the research;

4. Low cost: online surveys are easy to set up on the existing platforms with

minimal costs to the researcher. No travel is required of the researcher;

5. Participant anonymity: since data are collected online, no personal information is

shared and participants are at ease answering questions from the comfort of their

personal space;

6. Data codification: online surveys simplify data coding and organisation due to

inbuilt online tools that speed up the process and minimize human errors in data

entry;

7. Data analysis: data analysis can be easily performed using integrated quantitative

tools and statistical packages for both descriptive and inferential analyses;

8. Reducing pandemic risks: online surveys comply with the social distancing and

lockdown measures during the COVID-19 pandemic.

4.7.1 Data Collection Process

Figure 22 visualises the quantitative data collection approach used in this study. A

survey questionnaire was developed as the primary tool for data collection. A survey

questionnaire enables information to be gathered from a large number of IT

professionals in Saudi higher education institutes. For instance, positivist researchers

believe that the most appropriate research method is an extensive sample survey as it

provides a specific degree of control over the collection and analysis of data using the

parameters and statistical processes for the research design (Orlikowski & Baroudi,

1991). The items in the questionnaire reflect the constructs proposed in the conceptual

framework of the study. A focus group comprising a small sample of Saudi school

administrators was held to analyse their blockchain technology awareness. They

demonstrated limited knowledge of the technology which was attributed to the fact that

blockchain has only recently emerged and its practical applications in higher education

are not widespread. It was decided that to ensure the respondents did not misunderstand

any of the survey questions, a short description of blockchain technology would be

provided on each page of the questionnaire.

Figure 22: Quantitative Data Collection Process

4.7.2 Instrument Development Process

The questionnaire development procedure proposed by Moore & Benbasat (1991) is

used in this study. According to this process, questionnaire development takes place in

three stages: item creation, scale development and instrument testing.

•Stage 1: Item creation

The items for the questionnaire are developed to correspond to the constructs identified

in the finalised research framework. The research framework is finalised following the

completion of Phase I of the current research. Accordingly, the instrument, along with

the sources of items, is presented in Section 6.5. An important consideration of the

questionnaire preparation at this stage was the appropriate translation of the items.

Indeed, even though a questionnaire may have been validated in the country of origin, it

could only be considered robust in a different socio-cultural context if it maintains the

psychometric properties in the new language (Maindal, Kayser, & Norgaard, 2016;

Tsounis & Sarafis, 2018). In order to ensure a rigorous approach to questionnaire

development, professional translator services were used. The study employed the

forward-backward translation approach (Brislin, 1970; Jones, Lee, Phillips, Zhang, &

Jaceldo, 2001). First, a bilingual native Arabic speaker translator translated the

questionnaire into Arabic. Next, a different translator used the translated version to

backtranslate into English without looking at the original questionnaire version. The two

versions were compared to note any differences which would be corrected for the final

version upon agreement from the translators.

•Stage 2: Scale development

All items corresponding to the study constructs are based on a 7-point Likert scale to

maintain integrity and continuity. A 7-point scale is considered to be superior to a 5-

point scale for several reasons. Nunnally (1967), for example, determined that a 7-point

scale offers a better balance between the number of discriminant points and the

reliability of the items. Diefanbach et al. (1993) found that a 7-point scale is better

suited to reflect the subjective evaluations of respondents, such as beliefs regarding

technology applications. They reported that from the participant’s perspective, 7-point

scales were the most accurate and easiest to use. Sierles (2003) showed that data based

on a 7-point scale are more suited for advanced statistical analyses. Finally, Lewis

(2003) found that 7-point Likert scales correlated better than 5-point scales with the

significance levels. Based on these assertions, a 7-point scale was chosen with the

following answer options and coding: “1-strongly disagree,” “2-disagree,” “3-somewhat

disagree,” “4-neither agree nor disagree,” “5-somewhat agree,” “6-agree,” and “7-

strongly agree.”

•Stage 3: Instrument testing

In the last stage defined by Moore and Benbasat (1991), the questionnaire is pre-tested

before being distributed to the participants. In this study, pre-testing was based on the

independent evaluations of two groups. The first group consisted of IT professionals

who assessed the clarity of the questions and the other group comprised PhD students

who were asked to check if there were any issues related to the time needed to complete

the questionnaire and how easy the survey tool was to use.

4.7.3 Population and Sample

The target population of the study is the administrative and IT staff of higher education

institutions in Saudi Arabia. At the time of the study, there were 80 accredited colleges

and universities in the country, 14 of which were private. While it is difficult to

determine the exact number of administrative and IT specialists, a reasonable estimation

based on the researcher’s information on the two universities is that there were between

800 and 1000 individuals who met the criteria for participation in the study. The study

participants are individuals who held sufficiently high positions to enable them to assist

in formulating an institution’s technology development strategy, such as administrative

specialists (members of the boards, presidents, vice presidents, and deans) and

technology specialists (IT managers, technology consultants, CTOs).

In view of its relatively small size, the study aims to cover the entire target population.

Following Green (Green, 1991), the minimum number of participants is 104 + n, where

n is the number of independent variables: this brings the required minimum number of

participants to 115. Based on Nunnally and Bernstein (1967), the minimum number of

respondents is 10*n, which is 110. A more rigorous approach is to use a statistically

verified sampling size formula (Dattalo, 2008):

𝑋

𝑛 = 𝑁 ∗(1)

𝑋 + 𝑁 − 1

where

n is the minimum sample size; N is population

size; and X is a measure of confidence

estimated as:

𝑧! ∗ 𝑝 ∗ (1 − 𝑝)

𝑋

𝜀

z is a z-score; p is standard

deviation; ε is the desired

margin of error.

Using the formulas above, a sample of 115 participants from a population of 1000 has a

margin of error rate at 8.60%. To reach a rigorous 5% margin of error given the

population size, at least 278 responses are needed. Therefore, the study aims to collect at

least 278 responses for the maximum size effect, although anything above 110

participants is deemed acceptable.

4.7.4 Data Collection Process

The pre-tested, finalised, and ethics-approved questionnaire was placed online on the

Qualtrics survey tool which supports both the English and Arabic languages. To obtain

permission to reach the target population, letters of request were emailed to the college

and university administrators in Saudi Arabia. Their email addresses were obtained from

the institutions’ websites. The email contained the following: 1) a cover letter briefly

explaining the research and asking for assistance in recruiting study participants; 2) the

researchers’ requisites for further communication; 3) an attached approval form from the

UTS ethics committee. If a positive response was received, the administration was given

the survey link to be distributed among the potential study participants. If the

administration did not respond, a second request email was sent three days after the first.

If an institution declined to participate, no further action was taken.

4.7.5 Data Analysis

Figure 23 illustrates the quantitative data analysis process applied in this study. The data

analysis for the study is conducted with SPSS 22.0 and AMOS. The first part of the

analysis presents the sample description based on the answers from the General

Information section. The inferential analysis is conducted in two steps. First, a

measurement model is estimated for validity using confirmatory factor analysis (CFA)

(Thompson, 2004). The measurement model is used to test the validity and reliability of

the collected data. The internal consistency is tested with the Cronbach’s alpha at a 0.05

level of significance (Connelly, 2011). Discriminant validity analysis is performed using

square root average variance extracted (AVE) (Fornell & Larcker, 1981). Finally, a

common method bias test is performed using Harman’s single factor analysis

(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).

In the second step, the structural model is specified to examine the cause-and-effect

relationships between the study variables. The model’s fit is tested using a number of

parameters: the comparative fit index (CFI), root mean square error of approximation

(RMSEA), root mean square residual (SRMR), and Tucker–Lewis Index (TLI).

Following Kline (2015), the required CFI and TLI level is established as 0.9 or higher,

while both RMSEA and SRMR are set at 0.1 or lower. Additionally, scales were

adjusted and items were removed to achieve a good model fit if necessary.

The hypotheses were analysed through path analysis using the finalised structural model.

All tests were performed at a 0.05 level of significance.

4.8 Ethical Approval

Before carrying out the research activities, the relevant ethics forms were submitted by

the researcher to the UTS Human Research Ethics Committee at UTS for approval. The

participants were also informed by the researcher that at all stages of the research, the

code of ethics of UTS would be adhered to. The researcher also sent a formal request to

the Saudi Ministry of Education to carry out the research. Written information regarding

the aims and objectives of the study was given to all the participants based on language

preferences: in English or in Arabic (Appendix B). It was clearly stated in the consent

form that participation in this study is voluntary and that participation can be withdrawn

at any time. The consent form was available in either English or Arabic (Appendix C).

The process of data collection took place in a highly confidential manner and the

participants’ identities were kept anonymous. The data were only utilized for the

purpose of this research and the findings were limited to academic publications only.

4.9 Chapter Summary

This chapter described the solution proposed to address the research questions and the

chosen methodology. Following the science design approach, the chapter outlined a

sixstep process for developing the solution and addressing the major research question

and the associated subquestions. The solution starts with a review of theories that could

form the foundation for the model of blockchain adoption. Due to the focus of the study

on organisations and an intent to introduce additional contextual variables, DOI

(Rogers, 1995) and TOE (Tornatzky & Fleishcer, 1990) were selected to form the

theoretical foundation. These are comprehensive, validated organisation-level theories

that also offer due flexibility allowing the addition or removal of contextual variables.

These were supplemented with a quality dimension (Harvey, 2007) and barriers to

adoption (Clohessy, Treiblmaier, Acton, & Rogers, 2020).

To test and validate the proposed model, a mixed research methodology was selected

and justified. In Phase I of the model validation, it is presented to a group of industry

professionals for the initial valuation through an interview process. The final product is

the refined model with some original factors removed or added. In Phase II of the model

validation, a large-scale survey is conducted to determine the efficacy of the model in

predicting blockchain adoption in Saudi HEIs. This chapter described the methods of

data collection and analysis for both research phases.

99 The next chapter describes the proposed model and the relationships among its

variables.

5 The Model of Blockchain Adoption by Saudi HEIs

5.1 Introduction

This chapter presents and describes the initial model of blockchain adoption by Saudi

HEIs. The model was developed after a thorough review of the theoretical and empirical

literature on blockchain adoption. Section 5.2 visualises the model which is a

threedimensional DOI-TOE framework supplemented with education quality context

and a separate barriers dimension for practical purpose. Sections 5.2.1-5.2.5

operationalise and describe the role of each model dimension in the following order:

technology, organisation, environment, quality and barriers. For each dimension, an

explanation of the hypothetical role of each factor in the blockchain adoption process is

given.

5.2 Research Model

Taking the enhanced DOI-TOE framework as a basis, this study proposes that the

adoption of blockchain in Saudi HEIs is a function of five dimensions: technology

context, organisational context, environmental context, quality context and barriers to

adoption. The original framework is presented in Figure 24. The proposed framework

identified 21 factor relationships. The inclusion of the potential adoption factors in each

dimension was based on the review and analysis of the empirical literature and the key

postulates within DOI and TOE theories applied to blockchain technology (Chapters 2

and 4). Accordingly, the operationalisation and effect of each construct on blockchain

adoption in Saudi HEIs is discussed below.

Figure 24: Original Blockchain Adoption Model by Saudi HEIs

5.2.1 Technological Context

Technological features are important for the adoption process. Visible technologies that

can demonstrate value in excess of associated costs of implementation have a better

chance of adoption. To investigate the effect of the technological context on blockchain

adoption in Saudi HEIs, this study considered the impact of five technology attributes

described within DOI (Rogers, 1995).

Relative advantage refers to the degree to which an innovation is believed to provide

more benefits for an organisation (Rogers, 1995). In this study, relative advantage can

be defined as the degree to which blockchain technology is believed to be better than the

existing technology systems to administer educational process and offer educational

Technology Context

+Relative advantage

+Compatibility

+Trialability

+Observability

-Complexity

Organisational

Context

+Top management

support

+Organisational

readiness

+Organisation size

Environmental

Context

-Regulations

+Government

support

+Peer pressure

Quality Context

+Reducing

unemployment

+Education service

improvements

+Administration

improvements

Barriers

-Lack of knowledge

-Privacy and security

concerns

-Risk avoidance

-Lack of infrastructure

Lack of finance-

-Lack of specialists

-Lack of visibility

Blockchain

Adoption by

Saudi HEIs

services for higher education institutes. Within DOI, relative advantage arises from

weighing the innovation’s perceived benefits against the expected costs of adoption,

which are not necessarily financial in nature (Rogers, 2005). Different industries and

organisations recognise their own, specific advantages of this technology based on their

needs (Carson, Romanelli, Walsh, & Zhumaev, 2018; Clohessy & Acton, 2019).

However, the prerequisite costs of blockchain adoption for HEIs may also be substantial

given its relatively low level of maturity, scarce know-how and the lack of business

cases (Lustenberger, Malešević, & Spychiger, 2021). Given this, the adoption of

blockchain by a HEI will more likely occur if the institution considers it more

advantageous to the existing technologies in use. Previous studies demonstrated the

importance of perceived technology relative advantage for higher education institutions

(Abdekhoda, Gholami, & Vahideh, 2018; Tarhini, Al-Badi, Al-Gharbi, & AlHinai,

2018). Moreover, the relative advantage of blockchain over other technologies has been

empirically confirmed in a number of studies, including those conducted in the higher

education sector (Guo & Liang, 2016; Hartley, Sawaya, & Dobrzykowski, 2021; Iansity

& Lakhani, 2017; Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020). Therefore,

the following hypothesis is formulated:

H1a: Relative advantage has a positive influence on blockchain adoption in Saudi higher

education institutions.

The second factor within the technological context is complexity which is the firm's

perceived difficulty in understanding and using an innovation (Tornatzky & Fleischer,

1990). To be adopted, an innovative technology should be perceived as easy to

implement and use. It follows then, that the technologies which are considered

challenging and complicated have lower chances for adoption. By nature, blockchain

technology heavily relies on algorithms and cryptography which could be perceived as

complex by many. This brings uncertainty to potential adopters and a desire to wait and

understand better how it works (Clohessy & Acton, 2019; Drescher, 2017). Therefore,

perceived complexity poses a serious challenge to blockchain diffusion in organisations.

In fact, complexity is one of the most commonly mentioned factors related to

blockchain adoption with its negative impact clearly defined (Choi, Chung, Seyha, &

Young, 2020; Duan, Zhang, Gong, Brown, & Li, 2020; Malik, Chadhar, Vatanasakdakul,

& Chetty, 2021; Wamba, Queiroz, & Trinchera, 2020). Therefore, the following

hypothesis is formulated:

H1b: Complexity has a negative influence on blockchain adoption in Saudi higher education

institutions.

Trialability within DOI is defined as the extent to which an innovation could be tried on

a small scale before wider implementation (Rogers, 2003). Logically, if organisations

have an opportunity to try new technologies at little to no cost and resource

requirements, they are more likely to observe its benefits and decide to adopt. DOI also

posits that technology maturity plays a role: for organisations considering themselves

innovators and early adopters, trialability is essential (Rogers, 2003). It has been

previously demonstrated that blockchain is a rather immature technology in the higher

education sector, which many see as a main obstacle because organisations postpone

adoption until the proof of concept is visible (Schmitt, Mladenow, Strauss, &

Schaffhauser-Linzatti, 2019). At the same time, a series of case studies performed by

Clohessy and Acton (2019) demonstrated that organisations which were able to

experiment with blockchain relatively effortlessly on the cloud were more likely to

adopt it. Similar results were reported by Iansity and Lakhani (2017). Therefore, a

higher perception of trialability can play a positive role in its adoption. The following

hypothesis is formulated:

H1c: Trialability has a positive influence on blockchain adoption in Saudi higher education

institutions.

Observability within DOI refers to the level of technology visibility, recognition and

dissemination (Rogers, 2003). The theory proposed that technologies with tangible,

detectible attributes have a better chance of faster adoption. Contemporary blockchain

researchers have noted difficulties related to the observability of its effects (Dobrovnik,

Herold, Fürst, & Kummer, 2018; Lustenberger, Malešević, & Spychiger, 2021).

According to Rauchs et al. (2019), the main issue with blockchain applications is that

their tangible benefits are only observed after a while. Because of this, many

blockchainrelated projects are rejected without an opportunity to demonstrate the

results. Given this, a higher level of blockchain observability should increase the

likelihood of its adoption.

Therefore, the following hypothesis is formulated:

H1d: Observability has a positive influence on blockchain adoption in Saudi higher education

institutions.

The final technology context factor in the framework is compatibility. Within DOI,

Compatibility refers to the degree to which a new innovation is perceived as congruent

with the values, previous experiences, and requirements of potential adopters (Rogers

1995). Because these values and needs differ across industries and organisations, the

compatibility feature is important in the adoption process. When it comes to blockchain

adoption, researchers identified concerns related to its incompatibility with the existing

technical requirements and requirements related to data protection and management

(Lustenberger et al., 2021; Rauchs et al., 2019). There are also concerns related to the

absence of universal blockchain standards and, as a result, the availability of many

blockchain solutions which may not be fully compatible with each other and existing

organisational IT infrastructures (Holotiuk, Pisani, & Moormann, 2018). A number of

researchers have lately proposed that blockchain technology could benefit greatly if it

became perceived as compatible with organisational infrastructures and requirements

(Hartley, Sawaya, & Dobrzykowski, 2021; Kouhizadeh, Saberi, & Sarkis, 2021;

Morabito, 2017). Therefore, the following hypothesis is formulated:

H1e: Compatibility has a positive influence on blockchain adoption in Saudi higher education

institutions.

5.2.2 Organisational Context

While technology features are important for its adoption, the adopters themselves should

be ready for the process. Indeed, the integration of innovations with the existing

organisational processes, structures and strategy represents a challenging task for

management teams (Lalic & Marjanovic, 2010). Therefore, the successful adoption of

new technology requires the presence of key success factors within the organisational

context. Organisational context factors in this study are adopted from the TOE theory

and framework. According to TOE, organisational context includes organisational

resources, operations and features that play roles in innovation adoption (Tornatzky &

Fleischer, 1990). Top management support, organisational readiness and organisation

size are considered the key determinants of blockchain adoption in this study.

Top management support is the degree to which senior management has a positive

attitude towards a technology and its envisioning as contributory to organisational

success (Tashkandi & Al-Jabri, 2015). It is generally seen as a major factor for success

of all types of innovation projects in organisations (Chen, Xu, Lu, & Chen, 2018;

Crosby, Nachiappan Pattanayak, Verma, & Kalyanaraman, 2016). Top managers

promote the quality and effectiveness of innovative solutions by overcoming possible

resistance within their organisation, creating a vision for the innovation and providing

the required resources (Dong, Neufeld, & Higgins, 2009). Further, top management

support ensures adequate adoption planning and execution (Lustenberger, Malešević, &

Spychiger, 2021). This factor has received much attention in the blockchain adoption

studies as well. A number of researchers found that higher levels of top management

support are associated with higher likelihood of its adoption (Clohessy & Acton, 2019;

Hartley, Sawaya, & Dobrzykowski, 2021; Iansity & Lakhani, 2017; Wamba, Queiroz, &

Trinchera, 2020). Therefore, the following hypothesis is formulated:

H2a: Top management support has a positive influence on blockchain adoption in Saudi higher

education institutions.

Within TOE, organisational readiness is a comprehensive variable that includes different

aspects of preparedness to adopt innovations (Tornatzky & Fleischer, 1990). It can be

envisioned in the form of adaptiveness of the culture, structures and processes to new

technology adoption or in form of possessing sufficient resources for adoption (Iacovou,

Benbasat, & Dexter, 1995; Weiner, 2009). These resources comprise finance,

experience, knowledge, expertise and technology, all of which are allocated to the

innovation in question (Lustenberger, Malešević, & Spychiger, 2021). Blockchain, being

a relatively young, immature technology with a high degree of uncertainty, definitely

requires a higher level of organisational readiness. In fact, organisational readiness is

considered the top organisational context factor of blockchain adoption per mentions in

the literature (Clohessy, Treiblmaier, Acton, & Rogers, 2020). Available studies

confirmed the positive role of organisational readiness in the blockchain adoption

process (Clohessy & Acton, 2019; Crosby, Nachiappan Pattanayak, Verma, &

Kalyanaraman, 2016; Kouhizadeh, Saberi, & Sarkis, 2021; Zheng, Xie, Dai, Chen, &

Wang, 2018). Therefore, the following hypothesis is formulated:

H2b: Higher education institution readiness has a positive influence on blockchain adoption in

Saudi higher education institutions.

The final factor in the organisational context influencing blockchain adoption is

organisational size. It can be defined in terms of controlled assets, workers, market

share, total sales or other meaningful indicators (Bose & Luo, 2011). Within TOE, it is

expected that larger organisations are better suited to adopting innovations, especially at

the early stages of the diffusion process (Baker, 2012; Tornatzky & Fleischer, 1990).

This is because they have a larger amount of resources, a broader knowledge base,

higher levels of investment ability, and more developed infrastructures (Lee & Xia,

2006; Wang, Chen, & Xu, 2016; Zheng, Xie, Dai, Chen, & Wang, 2018). Some studies

questioned this postulate by arguing that smaller organisations are more capable of

adopting innovations due to agility, flexibility and control over resources and decision

making (Clohessy, Acton, & Morgan, 2017; Post, Smit, & Zoet, 2018). In relation to

blockchain, however, the majority of recent studies confirm the assertion that larger

organisational size is positively associated with adoption (Barnes & Xiao, 2019;

Clohessy & Acton, 2019; Hartley, Sawaya, & Dobrzykowski, 2021; Pilkington, 2016).

In this study, organisational size is defined by the number of students, which in Saudi

Arabia also reflects the amount of financial, human, and technological resources

available. Accordingly, a larger size is expected to contribute to blockchain adoption.

Therefore, the following hypothesis is formulated:

H2c: Higher education institution size has a positive influence on blockchain adoption in

Saudi higher education institutions.

5.2.3 Environmental Context

Environmental context refers to the external factors that influence the adoption of

innovation in an institution (Tornatzky & Fleischer, 1990). These factors are important

for two reasons. First, they are capable of influencing an organisation’s decision to adopt

innovation by exerting outside pressure; second, they are usually beyond the

organisation’s control and, therefore, cannot be effectively countered. Based on TOE,

three environmental context factors are considered influential to blockchain adoption in

this study: existing regulations, government support and peer pressure.

Emerging technologies which bring discontinuous changes to industries often represent

challenges for governments and legal systems because of uncertainty as to how to align

them with the existing laws and regulations (Piscini, Cotteleer, & Holdowsky, 2018).

Accordingly, it becomes difficult to forecast what the state of the regulatory

environment will be as the innovation in question becomes more widespread

(Lustenberger, Malešević, & Spychiger, 2021). For this reason, regulatory environment

is often seen as a negative factor for innovative technology adoption. It is well known

that blockchain today still lacks a definitive regulatory framework, Saudi Arabia not

being an exception in this case. The existing legal concerns are related to blockchain’s

distributive nature, data protection and taxation among others (Salmon & Myers, 2019).

As a result, analyses of blockchain theoretical applications and use cases often mention

regulatory uncertainty as an impediment to its adoption (Guo & Liang, 2016; Hackius &

Petersen, 2017; Kouhizadeh, Saberi, & Sarkis, 2021). It is, therefore, logical to assume

that the stronger the regulatory uncertainty regarding blockchain is, the stronger the

negative impact on adoption will be. The following hypothesis is formulated:

H3a: Existing regulations have a negative influence on blockchain adoption in Saudi higher

education institutions.

Government support can generally be thought of as a series of actions of various

character directed at stimulating technology adoption. This support can come in

different forms, including legal changes, subsidising innovations or mandating them

(Barnes & Xiao, 2019; Farooque, Jain, Zhang, & Li, 2020). Typically, government

support serves as a strong factor in technology adoption (Barnes & Xiao, 2019;

Clohessy & Acton, 2019). Given the novel nature of blockchain and the uncertainty that

many organisations assign to it, government support could mitigate some of the risks

associated with its adoption at the early stage of the diffusion process. This was

discussed in a number of papers on blockchain adoption where a stronger degree of

government support was seen as a desirable factor (Chen, Xu, Lu, & Chen, 2018; Choi,

Chung, Seyha, & Young, 2020; Malik, Chadhar, Vatanasakdakul, & Chetty, 2021;

Morabito, 2017). Therefore, it is expected that government support will be associated

with stronger blockchain adoption. The following hypothesis is formulated:

H3b: Government support has a positive influence on blockchain adoption in Saudi higher

education institutions.

The final environmental context factor is peer pressure. In general terms, peer pressure

is the influence exerted by partners, competitors and other organisations to adopt an

innovation (Sarkis, González-Torre, & Adenso-Diaz, 2010). Partners can pressure an

organisation to adopt novel technologies for stronger integration or to achieve

compatibility with more up-to-date systems (Iacovou, Benbasat, & Dexter, 1995). This

is especially true where partners exert a dominant relationship and there is a higher level

of dependency on them. In this case, technology adoption is more likely as the

organisation will try to maintain the partnership. At the same time, a partner’s support

may serve as an additional factor in fostering adoption (Hartley, Sawaya, &

Dobrzykowski, 2021). On the other hand, competitors may exert pressure to adopt an

innovation if they implement it en masse and/or there are visible positive results of

adoption. In this case, the innovation may quickly become a norm for a particular

industry (Chen, Xu, Lu, & Chen, 2018). Researchers of blockchain adoption in different

industries acknowledged the role of peer pressure in the form of partners’ pressure,

industry pressure and competitive pressure (Barnes & Xiao, 2019; Crosby, Nachiappan

Pattanayak, Verma, & Kalyanaraman, 2016; Kouhizadeh, Saberi, & Sarkis, 2021;

Lustenberger, Malešević, & Spychiger, 2021). Therefore, it is expected that peer

pressure stimulates blockchain adoption. The following hypothesis is formulated:

H3c: Peer Pressure has a positive influence on blockchain adoption in Saudi higher education

institutions.

5.2.4 Quality Context

Quality context is a sector-specific context introduced in this study. Based on the

available literature on quality measures for HEIs and public data for the Saudi education

sector, this study proposes that the quality context includes the following variables: 1)

intent to reduce graduate’s unemployment; 2) intent to improve HEI administration

efficiencies; and 3) intent to improve HEI service quality.

Improving access to the labour market and empowering students are considered key

indicators of education quality (Al-Ramahi & Odeh, 2020; Harvey & Green, 1993).

Unemployment reduction is one of the key issues for Saudi higher education to solve.

According to the latest data, youth unemployment (individuals under 24 years) in Saudi

Arabia is about 28%, and half of these individuals possess at least a bachelor’s degree

(O'Neill, 2022). As an innovation, blockchain has the potential to address this issue at

least partially. Two mechanisms could make this possible. First, blockchain offers an

accelerated, easily verifiable system of learning credentials and certificates (Abreu,

Coutinho, & Bezerra, 2020; Cheng, Lu, Xiang, & Song, 2020; Saleh, Ghazali, & Rana,

2020). Such systems make it easy for employers to verify students’ skills and education,

prevent fraud and help match the best candidates with open positions (Balon,

Kalinowski, & Paprocka, 2020). It also supports students’ career decisions and

personalised career recommendations based on learning achievements (Alammary,

Alhazmi, Almasri, & Gillani, 2019; Mikroyannidis, Domingue, Bachler, & Quick,

2018). Second, blockchain itself can create numerous job opportunities related to

engineering, software development, education, administration and related fields (Bucea-

Manea-Tonis, et al., 2021; Salah, Ahmed, & ElDashan, 2020; Shabaltina, Madiyarova,

& Tamer, 2021). Therefore, it is expected that intent to reduce graduates’ unemployment

will stimulate blockchain adoption. The following hypothesis is formulated:

H4a: Intent to reduce the graduate unemployment rate has a positive influence on blockchain

adoption in Saudi higher education institutions.

As discussed in Section 4.4.3 , education quality is envisioned through five components:

excellence, perfection, value for money, fitness for purpose and institutional outcomes

(Harvey, 2006; Harvey, 2007; Harvey & Knight, 1996). Innovative technologies can

serve to enhance one or more of these components thereby improving a HEI’s

competitive position. In fact, the adoption of innovative technologies is seen as a point

of differentiation and competitive advantage in the education sector (Tsinidou,

Gerogiannis, & Fitsilis, 2010; Waller, Lemoine, Mense, & Richardson, 2019). Two

mechanisms for this are possible: 1) HEIs can improve administration efficiencies to

reduce the cost of operations (and possibly tuition cost) and 2) HEIs can improve

service quality by offering novel approaches to education and the products of education.

There is evidence that blockchain technology can serve both purposes. In terms of

operational efficiencies, blockchain has shown very promising results due to its inherent

features allowing intermediaries to be eliminated (Alammary, Alhazmi, Almasri, &

Gillani, 2019; Awaji, Solaiman, & Albshri, 2020; Ma & Fang, 2020). Likewise, there is

also growing evidence that blockchain adoption has brought service improvements in

healthcare, supply chains, and other sectors (Loizou, Karastoyanova, & Schizas, 2019;

Tandon, Dhir, Najmul Islam, & Mäntymäki, 2020; Tijan, Aksentievic, Ivanic, & Jardas,

2019). Therefore, it is expected that HEIs’ intent to improve administration efficiencies

and service quality will serve as stimulating factors for blockchain adoption. The

following hypotheses are formulated:

H4b: Intent to improve administration quality has a positive impact on blockchain adoption by

Saudi HEIs.

H4c: Intent to improve service quality has a positive impact on blockchain adoption by Saudi

HEIs

5.2.5 Barriers to Adoption

Despite its potential benefits for higher education, blockchain has yet to receive

widespread adoption in the industry (Alammary, Alhazmi, Almasri, & Gillani, 2019;

Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020). While it can be argued that the

technology still remains in its early stages of the diffusion process, it is also likely that a

series of factors may slow down its acceptance. Within DOI and TOE, some barriers to

adoption are already embedded in the respective frameworks: examples are technology

complexity and government regulations. However, research on blockchain applications

has identified more potential barriers within each of the key dimensions (Choi et al.,

2020; Clohessy et al., 2020). This study, therefore, considered the most commonly

mentioned barriers in the literature and put them into a separate factor dimension. This

was done for practical purposes: listing the relevant barriers separately allows decision

makers to devise specific steps to overcome them.

Knowledge about technology is an important feature that distinguishes adopter

categories within DOI (Rogers, 2003). Innovators and early adopters are the ones who

acquire knowledge about technology and its benefits faster and, therefore, are first to

adopt them. As knowledge spreads further, other categories join the process.

Consequently, a lack of knowledge is usually considered one of the main factors which

slows technology adoption, including technologies potentially beneficial to higher

education (Abrahams, 2010; O’Doherty, Dromey, & Lougheed, 2018). With regard to

blockchain, Alam (2022) argued that “many educational stakeholders are ignorant of the

benefits and possibilities of blockchain technology owing to a lack of knowledge of this

technology” (p. 307). Similarly, Choi et al. (2020) proposed that the lack of general and

technical knowledge about blockchain is one of the main organisational level factors of

adoption. The need to improve blockchain knowledge among organisations for faster

adoption was emphasised by several other researchers (Guo & Liang, 2016; O'Dair,

Beaven, Neilson, Osborne, & Pacifico, 2016; Tapscott & Tapscott, 2016). Therefore,

given both the theoretical and empirical evidence regarding the relationship between

knowledge about technology and its adoption, it is assumed that the lack of knowledge

about blockchain will be a barrier to its adoption. The following hypothesis is

formulated:

H5a: Lack of knowledge about blockchain has a negative impact on blockchain adoption

by Saudi HEIs.

Another major negative factor related to blockchain adoption is privacy and security

concerns. These are interrelated concerns focusing on who gets access to data and

whether there are sufficient mechanisms to protect the data from being used for

malicious purposes (Shin, 2019). Privacy and security concerns affect adoption decision

making at both individual and organisational levels in higher education institutions

(Becker, Newton, & Sawang, 2013; Qasem, Abdullah, Jusoh, Atan, & Asadi, 2019;

Singh & Hardaker, 2014). Being a “trustless” type of system, blockchain enables

transparent transactions between parties that do not have to know and/or trust each other

(Sillaber, Waltl, Treiblmaier, Gallersdörfer, & Felderer, 2021). Further, blockchain

system participants are able to review the previous transactions and their participants,

especially in public blockchains (Iansity & Lakhani, 2017). While some researchers and

practitioners believe that such level of distributed trust and transparency are beneficial,

others argue that it is a weakness (Kosba, Miller, Shi, Wen, & Papamanthou, 2016;

Underwood, 2016). Some authors specifically pointed to extremely high levels of

visibility in blockchain networks as a roadblock for many organisations to join the

adoption trend (Babich & Hilary, 2019; Choi, Chung, Seyha, & Young, 2020). Further,

some researchers pointed to a possibility of entering dubious or erroneous code into

smart contracts as a possible security concern due to the immutability of the system

(Surujnath, 2017). Finally, privacy and security concerns are found to be the top

individual factor which affects organisational decision makers when it comes to

blockchain (Clohessy, Treiblmaier, Acton, & Rogers, 2020). Therefore, the following

hypothesis is formulated:

H5b: Privacy and security concerns have a negative impact on blockchain adoption by Saudi

HEIs.

Risk is one of the factors associated with adopting innovations within DOI (Rogers,

2003). The earlier adopters are usually more prone to take risks while the later majority

category prefers to avoid risk at all costs. These are generalised observations, however,

as risk is treated differently by individuals and organisations (Prewett, Prescott, &

Phillips, 2020). At the organisational level, risk is a multi-layered concept which

involves various hazards: business risks, financial risks, operational risks, legal risks

being some. Typically, an organisation tries to minimise these hazards or their negative

possible effects, which is known as risk avoidance (Sun, 2021). Arguably, blockchain

technology carries a number of uncertainties for organisations related to

implementation, governance and benefits (Alammary, Alhazmi, Almasri, & Gillani,

2019). Accordingly, perceived risks emerge in many areas associated with blockchain

adoption and use. Moreover, blockchain dismantles the traditional notions of trust and

the associated third-party institutional support (Sadhya & Sadhya, 2018). Therefore, it is

expected that the desire of organisations to avoid risk will negatively influence

blockchain adoption. The following hypothesis is formulated:

H5c: Risk avoidance has a negative impact on blockchain adoption by Saudi HEIs.

For some organisations, the perceived benefits of blockchain adoption may still

outweigh the aforementioned concerns. However, even if an organisation is willing to

adopt a technology, it must possess adequate resources for its successful

implementation. Weiner (2009) proposed that if an organisation intends to successfully

adopt an innovation, at least three types of resources should be available: human

resources (knowledge and skills), financial resources (money and budget) and

infrastructural resources (technologies). When one or more of these factors is absent, the

change of successful adoption diminishes (Clohessy & Acton, 2019). There is a lack of

research on the specific influence of each of these factors on blockchain adoption in

education. However, some general evidence may offer insights. Babich et al. (2019)

found that resistance to blockchain adoption may arise from a lack of workers’

expertise. Likewise, Prewett et al. (2020) argued that without a sufficiently qualified

workforce, organisations will be unable to extract the full benefit from blockchain

applications. Further, a lack of appropriate IT infrastructure as an impediment to

blockchain adoption was emphasised by several authors (Iansity & Lakhani, 2017;

Lindman, Tuunainen, & Rossi, 2017; Swan, 2015). Finally, if an organisation lacks

finances, it may not be able to acquire either the necessary human resources or

technology to support successful blockchain adoption (Choi, Chung, Seyha, & Young,

2020; Hughes, et al., 2019). Therefore, it is expected that a lack of appropriate

technology infrastructure, finances and/or human specialists will be detrimental to

blockchain adoption. Accordingly, the following hypotheses are formulated:

H5d: Lack of IT infrastructure has a negative impact on blockchain adoption by Saudi HEIs.

H5e: Lack of financing has a negative impact on blockchain adoption by Saudi HEIs.

H5f: Lack of human resources has a negative impact on blockchain adoption by Saudi HEIs.

Finally, successful technology adoption may be derailed by the lack of its visibility.

According to DOI, technologies that are visible and have detectable benefits and

attributes have a higher chance of adoption (Rogers, 2003). This is because visibility

creates knowledge and increases confidence in the applicability and ability of

technology to meet the needs of potential adopters. Since blockchain is still in its early

stages of dissemination in higher education, its visibility may still be limited. Some

researchers specifically attributed the slower degree of blockchain adoption to the fact

that colleges and universities do not see how the technology works and what kind of

benefits it offers (Dobrovnik et al., 2018; Lustenberger et al., 2021). Rauchs et al. (2019)

suggested that blockchain diffusion is low because it takes time to recognise the real

benefits of its use. The lack of visibility due to the relatively low number of successful

business use cases of blockchain was recognised in a number of works (Iansity &

Lakhani, 2017; Kamble, Gunasekaran, & Arha, 2018; Zheng, Xie, Dai, Chen, & Wang,

2018). Choi et al. (2020) found that the lack of business examples of successful

blockchain application reduces management commitment and support. Therefore, it is

expected that the lack of blockchain visibility may adversely influence blockchain

adoption. The following hypothesis is formulated:

H5g: Lack of visibility has a negative impact on blockchain adoption by Saudi HEIs.

5.3 Chapter Summary

This chapter presented the original model of blockchain adoption by Saudi HEIs. The

model was developed on the basis of the theoretical and empirical literature on adoption

in higher education. Five dimensions and 21 factors influencing the adoption process

were identified, and their role was justified. Accordingly, hypotheses were formulated to

explore the significance and role of each factor and dimension as a whole on the

adoption of blockchain in Saudi HEIs.

The next chapter presents the results of the model presentation and analysis by a group

of technology and HEI administration experts.

6 Research PHASE 1: Model Presentation and Analysis

6.1 Introduction

This chapter offers a review of the qualitative data collection and analysis. The process

involved a series of interviews following the Blockchain Adoption Model presentation

to a number of industry experts. The main two goals of the qualitative data analysis

were to single out the most promising areas for blockchain applications in Saudi higher

education institutions (HEI) and to determine the most important factors of adoption

according to the decision makers. Section 6.2 describes the interview sample, that is, the

individuals who participated in the study. Section 6.3 describes the results of the

interviews: 1) the participants’ views on the most viable areas of blockchain application

in higher education and 2) an analysis of the factors that could be influential in the

blockchain adoption process by Saudi HEIs. Section 6.4 presents the refined model of

blockchain adoption following the interview results. The refined model includes the new

factors proposed by the interviewees and removes some of the factors that the

interviewees did not consider important. Finally, Section 6.5 presents the final version of

the questionnaire to evaluate the model on a large population. The questionnaire

includes the items corresponding to the finalised set of factors after the interviews.

Sections of this chapter have earlier been published in the following conferences

articles:

Alalyan, M.S., Jaafari, N.A., Hussain, F.K. (2023). Technology factors influencing Saudi

higher education institutions’ adoption of blockchain technology: A qualitative

study. Advanced Information Networking and Applications (AINA). Cham:

Springer International Publishing, pp. 197-207.

Alalyan, M. S., Jaafari, N. A., & Hussain, F. K. (2023). Barriers to blockchain adoption

by Saudi higher education institutions: A structural equation analysis. Advances

in Networked- based information systems., In press.

6.2 Data Collection Results and Sample Description

In total, 10 interviews were conducted with the mid-level and senior-level administrative

and information technology (IT) specialists from Saudi HEIs. Following the general

recommendations in the literature (Brinkmann & Kvale, 2015; Weller, et al., 2018), a

data saturation approach was applied. No substantially new insights on the topic were

obtained and no new themes emerged during the tenth interview, after which it was

concluded that the data saturation point was reached. The collected data were deemed

sufficient for qualitative analysis.

The sample of the interviewed individuals is displayed in Table 20.The sample was

balanced in terms of position and type of institution. There were three high-level

administrative personnel interviewees and seven information technology specialists

representing the middle and high levels of responsibility. Eight individuals in the sample

represented public universities, and two represented private universities. This proportion

approximately corresponded to the general distribution of private and public HEIs in

Saudi Arabia as discussed in Chapter 2. Five interviewees described their level of

knowledge about blockchain as adequate, average, or familiar while the remaining

described it as good or excellent. As such, the participants possessed a sufficient level of

blockchain expertise for the purpose of the study.

Table 20: Interview Sample

N Name Position Institution Institution

Type

Level of

Blockchain

Knowledge

1 P1 IT Support Department of

Education Ministry Some

2 P2 IT Deanship Web

Developer

King Khalid

University Public Some

3 P3 IT Deanship Islamic University

of Riyadh Public Good

4 P4 Technical Support King Khalid

University Public Some

5 P5 Head of Computer

Science Department Taif University Public Good

6 P6 IT Administrator KAUST Private Some

7 P7 Service Desk Manager King Khalid

University Public Some

8 P8 Dean of Computer

Science and Engineering Hail University Public Good

9 P9 IT Project Manager KAUST Private Good

10 P10 e-Learning Specialist King Khalid

University Public Good

6.3 Data Analysis

6.3.1 Basic Approach

As discussed in Chapter 4, qualitative data analysis followed the approach developed by

Miles and Huberman (2019). The first step in the qualitative data analysis involved data

reduction and organisation, which was followed by pattern coding and, finally, data

display and interpretation. To better contribute to the study goals and to meet the needs

of qualitative data collection and analysis, two types of analysis are presented below.

The first part describes the interviewees’ perceptions regarding potential blockchain

benefits and areas of application for Saudi HEIs. The second part involves descriptive

conceptualisations of independent dimensions and constructs related to blockchain

adoption as well as identifying the potential links among these constructs in the process

of blockchain adoption.

6.3.2 Blockchain Benefits and Potential Applications

At the onset of the interview, the participants were asked about their thoughts regarding

blockchain use in education globally and Saudi Arabia in particular. In total, 8 out of 10

interviewees were ready to speak on this subject. The general consensus among the

interviewees was that blockchain is just starting to make its way into the education area.

The adjectives used to describe the adoption level were “generally low” (4

interviewees), “some” or “partial” (2 interviewees) and “in its infancy” (2 interviewees).

P3 suggested that the level of adoption is “probably higher in the developed nations”

whereas P8 proposed that public universities would be less prone to adopt blockchain

because of “a somewhat cautious bureaucracy and waiting for some results and to

observe some applications at other universities.” She saw it as one of the main reasons

that Saudi HEIs are behind in terms of blockchain adoption in comparison to other

countries’ HEIs. In relation to Saudi HEIs, 4 interviewees had no knowledge about any

blockchain uses, whereas 2 interviewees thought that blockchain could be used to some

extent, but they were not sure where or how. Only 4 respondents were able to

confidently speak about blockchain use in Saudi HEIs, particularly about certificates at

KAUST.

Table 21 provides a list of codes developed from the analysed interview transcripts for

blockchain benefits and potential areas of application dimensions. Of the 10

interviewees, 8 spoke about the ways that Saudi HEIs could benefit from blockchain.

Five types of benefits were mentioned. The majority of interviewees (7 out of 8)

mentioned security improvements in relation to data, records and transactions (BEN-

SEC coding). Additionally, 6 out of 8 interviewees spoke about the benefits of faster,

more efficient transactions (BEN-EFF coding); improvements in data management, its

verification, retention and access (BEN-DAT coding); and the decentralisation of trust

with no need to rely on a third party for confirmation (BEN-DEC coding). Fewer

interviewees (3 out of 8) also mentioned the benefit of information exchange between

HEIs, employers, and students (BEN-EXC coding).

Table 21: The List of Codes for the Blockchain Benefits and Potential Areas of

Application Dimensions

Code Meaning Mentions

BEN Perceived Blockchain Benefits 8

BEN-SEC Security of data, records, transactions 7

BEN-EFF Efficiency of transactions: speed, time 6

BEN-DAT Data management improvements: verification,

retention, access

BEN-DEC Decentralisation of trust: absence of third-party

reliance necessity

BEN-EXC Information exchange, confirmation, approvals 3

ADP Application Areas 10

ADP-CER Certificates, diplomas 10

ADP-IPP Intellectual property protection 9

ADP-ADM Administrative procedures 8

ADP-EVA Student evaluations 7

ADP-ENH Enhanced learning applications 6

ADP-LRN Learning and assessments 4

ADP-SMT Smart environments 4

In the next step, the interviewees were asked to identify the potential areas of

application for blockchain in Saudi HEIs with the goal of determining the most feasible

application areas for further investigation. Based on the review of the available literature

on the theoretical and practical applications of blockchain in education, seven topics

were discussed. The results are reviewed below.

The most commonly mentioned application was in relation to certificates and diplomas,

which was discussed by all interviewees. Specifically, the interviewees mentioned such

areas as diploma issue, confirmation and immutability. In this regard, some notable

responses included2:

Firstly, to preserve the environment from the waste that happens... Also,

the diploma will be reserved at the student, non-perishable, not lost… It

solves many problems with blockchain certificates. (P3)

… there is a saving of time and very high security and maintaining the

certificates from forgery or impersonation, for example, and this is

possibly one of the most important things I mean in the certificates. (P4)

Blockchain is essential to apply in certificates because it helps students

deliver theirs anytime and from anywhere. (P5)

From my point of view now, I mean, it is considered the easiest, and the

first application always for universities, mainly for the blockchain, is on

certificates and accreditation of university certificates. Because it is a

straightforward application, the verification process is also much easier

than any other product and app. (P8)

2 Hereafter, direct quotes from the interview transcripts are provided. Some language errors are likely

because for the interviewees, English is not their native tongue.

Intellectual property protection was the second most commonly mentioned blockchain

application with 9 interviewees discussing it, although there was apparently less

confidence in how it could be used and to what effect. Those who clearly supported

blockchain use for academic intellectual property protection (5 interviewees) linked it

mostly to the idea of an overall blockchain-enhanced digital ID which also offers a

higher degree of protection:

When it is in ID and this advanced thought will work and will help that

the one means that there is protection, protection of data. (P1)

Every student should have a digital ID that allows the student to keep his

certificate, transcript, and research. In the future, this makes it easy for

the student to search and apply for a job very quickly. (P4)

I think it's a good technology because it has high security and has more

credibility in protecting researchers' rights. (P9)

At the same time, some interviewees expressed doubts about whether blockchain can

enhance intellectual property protection in academia because of other means of

protection:

Different methods can be used for intellectual property protection.

However, I feel it is still jurisprudence. I feel that it has not become more

transparent […] for the person who owns the idea. […] Even I have a

record registered with intellectual property. (P2)

In terms of research and scientific theses, they inevitably maintain that

when there is a quotation, it is a documented quotation, not without the

knowledge of its author. (P6)

Yet others thought that intellectual property protection is not on the blockchain agenda

of HEIs because they have other, more urgent priorities:

Intellectual property protection is not a priority to apply to blockchain

because, at the moment, there are many technologies used to protect

intellectual property. (P5)

It is not essential to me as a university official at the moment. However, at

the moment, I need university certificates, academic records, admission,

and the process of student admission and registration. (P8)

Yet another frequently mentioned area of application was school administration, which

was discussed by 8 of the 10 respondents. The interviewees discussed the promising

applications in the registration process (6 interviewees), admissions (4 interviewees),

data extraction and verification, grade confirmations and transfers (1 each). Blockchain

was seen as a technology to enhance transparency, improve student data security, and

reduce transaction times for data transfer, confirmation and verification. Some examples

of the responses are:

For example, if we have a student who wants to register at university,

suppose that we have converted electronic certificates and papers that I

need to accept and register electronically. Blockchain made it easier for

that student to securely apply once to all universities. (P3)

In student registration, blockchain means the easy extraction of data.

And verification data. If corruption occurs, all the data will not be

damaged or lost, especially when the student graduated a long time ago.

(P6) I give 10 for Administration, because it is a saving of time and a

transparency provision. (P8)

Blockchain uses for student evaluations were discussed by 7 interviewees. While 5 of

the interviewees expressed positive views regarding blockchain applications in this area,

only two were able to distinguish them from the already existing online technologies.

Specifically, features such as prevention from grade and achievement manipulation and

continuous skill and achievement update were discussed:

Blockchain helps to prevent the grades from being manipulated. The

grades and achievements cannot be changed.

There is a significant point: we can change and benefit from blockchain

with artificial intelligence in evaluating the student and in giving him or

developing his skills. This means, I can tell the student from the day he

enters the university that you will have a file of your skills. (P3)

Others referred to the already existing online features and technologies without making

specific references to blockchain:

Technology will make it easier for the faculty member to correct

assignments and tests online, publish the grades and make them easily

accessible for students. (P5)

Evaluating students with tests and assignments can contribute to

speeding up the correction process and thus reduce the correction burden

on the faculty member. (P9)

Because we currently have online student evaluations, the faculty will

enter the grades and evaluations at different stages of the learning

process to track and demonstrate the student progress. (P10)

One interviewee also expressed an opinion that blockchain is “not essential” for student

evaluations, and other, more urgent applications should be considered.

The interviewees also spoke about blockchain applications for enhanced learning

environments (6 interviewees in total). Of these, 4 considered blockchain as an

important tool given that education is moving increasingly online and new methods of

reliable, effective content delivery are in demand:

Based on the features of the blockchain, it is better than Blackboard.

Because Blackboard suffered a lot last year, you know the system is

frozen, continuous pressure, sure blockchain will work. (P1)

In the Covid-19 pandemic, we needed all the educational environments to

become online, whether in quizzes or assignments. The exams help a lot

that when each student has a unique record in it from the beginning of

the study to the time of his graduation, he will have this experience that

he can refer to when he is looking for a job. (P2)

The whole world has carried technology with it in every place and uses

technology in every place. […]. For example, or you are working on the

project electronically. Some programs and applications help with this

thing. Moreover, I expect that even if they meet as a class as a form of

attendance physically. I expect that they will also use the technologies,

whether the Blockchain or the programs that help them record the

meeting (P3)

Others, however, expressed more cautious opinions. P8 stated that “there is no clear,

observable path for blockchain use in enhancing existing learning environments” at this

point in time. Similarly, P10 argued that “an increasing number of commercially

available enhanced learning systems are being offered in the distance education market”

thereby meaning that demand for blockchain applications may not be sufficient.

Only 4 interviewees spoke about blockchain’s potential applications for students’

learning and assessments. Of these, however, only 2 expressed confidence in

blockchain’s real contributions, mostly in terms of tracking students’ progress and

keeping educational achievements in one place:

Look at the Covid-19 pandemic. It showed us the need to check all

education achievements in one place. This differentiates blockchain from

other technologies. You can keep all learning records ever and make

them available for check in literally one click. (P3)

For those who keep educating themselves, through online courses or

different platforms, they would want to add information to one place,

update it, and then make it presentable and verifiable. This is where

blockchain becomes very useful. (P1)

The remaining 2 interviewees acknowledged blockchain had some potential for

verifiable CVs which could be updated and verified immediately after new education

milestones are achieved. At the same time, they noted that it is not something being

developed at the moment:

Take the job platforms where people post their CVs to find jobs.

Blockchain may ensure that these CVs are up to date and free from

misinformation. Maybe with time, they can come to it, but I see that it's

not ready right now. (P10)

Colleges and universities have other priorities when it comes to

blockchain. In the future, yes, you may see how students progress with

specific learning topics and include their learning achievements in

lifelong learning blockchain. However, it is not the foreseeable future as I

see it. There are other instruments available. (P8)

Finally, 4 interviewees discussed the use of blockchain to create smart learning

environments. However, such discussions mostly concentrated on distant future

applications rather than something expected in HEI soon. The interviewees

acknowledged the idea, but stated that it was too early to consider it seriously:

A university must first start applying blockchain technology in all its

electronic, administrative, and educational transactions. Then it will

reach the point of being a smart university. (P5)

At the current time, no, I see no feasible blockchain-enhanced smart

universities. However, in the long run, hopefully, it will be possible. At

present, people still have ignorance in blockchain technology and how to

apply it to the whole education process. (P1)

I cannot find examples of smart universities now, I mean, they are not

100% smart. […] However, if the trend emerges, and you are able to

quickly explore it and implement it, you will be very successful. (P3)

Based on the analysis of the interviewees’ perceptions of blockchain potential

applications in HEI, several important themes emerged:

Theme 1: blockchain is still an emergent topic among IT professionals and HEI

administrators in Saudi Arabia.

Of the 10 interviewees, only 5 expressed a deep understanding of the topic, whereas

another 5 acknowledged somewhat limited knowledge. There was no consistent

understanding among the interviewees regarding blockchain adoption in education,

including Saudi Arabia. Some were not able to tell in which particular areas blockchain

is being currently used and applied in education. With some exceptions, the interviewees

spoke of blockchain as a type of new technology which needs to be investigated and

researched further rather than something being readily adopted and applied.

Theme 2: the primary benefits of blockchain for education, as recognised by the

interviewees, are its decentralised nature and security trust mechanism.

The interviewees considered these aspects of blockchain as particularly useful in

managing student records and credentials and offering novel ways to enhance

administrative tasks and learning processes. Specifically, the mechanisms of security,

immutability and fast verification within the network were acknowledged. On the other

hand, not all interviewees were able to directly relate the particular aspects of

blockchain to novel applications in education. Sometimes, there was little clarity about

how exactly blockchain implementation would be different from the already existing

learning technologies.

Theme 3: the most promising area for blockchain applications in Saudi HEI is

blockchain certificates and diplomas.

During the interviews, the participants were asked to name the potential blockchain

application areas for higher education and rate them. A summary of these assessments is

presented in Table 22. As can be seen, blockchain-supported school certificates and

diplomas were the only higher education area mentioned by all interviewees and

consistently received the highest scores. Blockchain applications for administrative

tasks and academic intellectual property protection were other commonly mentioned

areas, although there was a certain amount of disagreement among the interviewees

regarding their potential and viability. Other potential areas of blockchain applications

were not discussed as much and/or lacked substantial consensus regarding benefits and

uses.

Based on the themes developed during the analysis, it was decided to proceed with the

focus on blockchain applications for university diplomas and certificates.

Table 22: Interviewees' Assessments of Blockchain Potential for Higher Education Aspects

Interviewee Certificates Administrative

Tasks

Learning

and

Assessment

Student

Evaluation

Enhanced

Learning

Protection

Smart

Environment

P1 10 10 10 8.5 10 10 -*

P2 10 - - - 10 8 -

P3 10 10 10 10 8.5 10 10

P4 10 8 - - yes** 10 yes

P5 10 8 - yes - - 5

P6 10 10 - 9 - 6 -

P7 10 10 - - - - -

P8 10 10 6 6 6 6 6

P9 10 - - 10 - 10 -

P10 10 10 8 10 10 10 -

Total

Evaluations 10 8 4 7 6 8 4

* ‘-‘ indicates that the interviewee did not provide any discussion on the topic **

‘yes’ indicates that the interviewee mentioned the area of application but did not

rate/evaluate its potential

6.3.3 Factors Influencing Blockchain Adoption in Saudi HEI

Based on the theoretical underpinnings of the study, the interviewees were asked about

the influence of multiple factors on blockchain adoption by Saudi HEIs. Additionally,

the interviewees were asked to offer their own input about any additional factors they

found significant in this process. As a result, six blocks of factors were discussed. The

following analysis of each block of factors is based on Miles and Huberman’s (2019)

approach where 1) data reduction was applied reduce the data into an analysable format;

2) pattern coding was applied to develop themes from the content analysis; and 3)

summaries of codes and underlying themes were presented in the form of tables for

visualisation and interpretation.

For the data reduction process, operational codes were developed to identify constructs

related to blockchain adoption. Additional coding was developed to identify the strength

of each considered factor in relation to blockchain adoption in HEI. Four types of codes

were developed in this regard:

- Strong effect value was assigned to constructs which were identified by the

interviewees as “strong,” “very important,” “important”, “essential,”

“extremely,”, “imperative” and similar adjectives.

- Moderate effect value was assigned to constructs which were discussed by

the interviewees using words “moderate,” “not very strong,” “sensible,”

“restrained,” “average” and similar terms.

- Unclear effect value was assigned to constructs which were discussed by

the interviewees using such terms as “unclear,” “not sure,” “maybe,”

“perhaps”, “probably,” “could be” and similar terms.

- Finally, no effect value was assigned to constructs whose influence was

clearly denied by the interviewees.

The analysis also demonstrates the mechanisms through which the aforementioned

effect values were assigned by the study participants.

6.3.3.1 Technological Factors

Data codes and the frequency with which technological factors were mentioned are

listed in Table 23. The coding is ranked based on factor strength. As can be seen,

complexity was the most discussed technology effect in the context of blockchain

adoption, although relative advantage by far was the most consistently mentioned and

the most highly rated. Overall, none of the theoretically established technology factors

was considered to be weak or irrelevant by the study participants.

Table 23: Coding Applied to Technology Dimension Factors

Strength of Effect on Adoption

Code Meaning N Strong Moderate None Unclear

TECH-RA Relative Advantage 7 7 - - -

TECH-

COMP

Compatibility 9 5 - 3 1

TECH-OBS Observability 8 2 2 2 2

TECH-CPX Complexity 10 3 3 3 1

TECH-TRIA Trialability 8 3 3 1 1

Several themes clearly emerged in relation to the effect of relative advantage on

blockchain adoption by HEI. Five interviewees considered it an “essential technology”

for future applications whereas the other two called it a “very important” aspect of

blockchain. Blockchain’s relative advantage was discussed in the context of catering to

the growing amount of education moving online, specific needs of HEIs in managing

the process, and overall being forward looking, value adding technology with next-level

security mechanisms:

Relative advantage is an essential factor. Because trust, privacy, and

security in blockchain are high. It is essential in terms of the policies that

we always work with cybersecurity and in any conflict. (P2)

Speaking of the relative advantage of blockchain, first of all, it can cater

to specific business needs. As an administrator, for example, I know what

the needs of my university are. Like, let me say, for example, my

University is a women's University. I am sure that our University's has

specific needs. So, I know first of all what I need in technology for

blockchain, and depending on the need, I specify where I can apply it.

(P7)

Relative advantage is an essential factor. Because sometimes, it is not a

requirement that any trendy technology will be compatible with existing

applications. Sometimes it's ok; it saves time and it offers

decentralization. (P8)

Blockchain also has the advantage of being distinguishable from the rest

of the technologies we have, and this feature may support us in its

application which is purely digital technology. This will be an added

value for the University, especially if one of the objectives of the

University is to be a digital university. (P9)

Given the high value assigned by the interviewees to the relative advantage factor in

blockchain adoption by HEI, it was decided to retain this factor in the final conceptual

framework for the study.

With regard to compatibility, the opinions of the study participants were divided. Of the

9 interviewees, 5 considered it a strong factor in the blockchain adoption process. They

spoke about blockchain’s compatibility with their institutions’ goals and objectives,

administrative procedures and educational services:

I think that in our case, in our university, blockchain is compatible with

the University, our vision, message, goals, and so on. (P1)

Compatibility is an essential factor. I mean, it generally is if it integrates

with the existing infrastructure in the administration. Is the

administration ready to integrate blockchain? Sure, in our case I think it

is. I think that this factor is essential: its compatibility with the university

infrastructure. (P2)

Compatibility is an essential factor. At least in my area of responsibility,

which is service provision: education services and supplemental services.

Blockchain should be compatible with what we do and, better, with the

existing systems we use. (P7)

Of course, compatibility matters, especially if this technology allows the

connection of other tools, I mean, connectibility to other different

systems. (P10)

On the other hand, some interviewees denied compatibility had a meaningful impact or

stated that it was much weaker in comparison to relative advantage:

I do not think that compatibility is necessary for applying blockchain

technology. Because blockchain technology can be applied and adapted

to match the environment in which it is applied. (P5)

With regards to compatibility… Well, in our case if our school uses a new

technology, so, it must not affect the old systems. It will probably replace

them outright. So, not that essential of a factor. (P6)

I think that incompatibility or compatibility is not an obstacle or things

that may affect the adoption of the technology. (P8)

Accordingly, the themes that emerged in the discussion of the compatibility factor in

blockchain adoption were: 1) compatibility is sought in all aspects of HEI management:

from services to administration thereby making compatibility an important factor in

blockchain adoption; 2) from a technical standpoint, compatibility with the existing

systems may not be that important for decision makers. Given the importance assigned

by the interviewees to compatability, it was decided to retain it for the final conceptual

model of the study. However, the influence of the factor in the final model would not be

expected to be as high as the relative advantage factor.

Blockchain observability in the adoption process was discussed by 9 interviewees. The

opinions regarding its effect were clearly divided. Some interviewees argued about the

strong influence of observability, especially in view of the positive effects in other

institutions or in some areas of their own institution:

If a new technology appears, we must keep pace with it, especially if it is

applied in a university. Moreover, suppose I noticed that this university

succeeded in using the technology, facilitated many tasks, and reduced

costs. In that case, it will encourage the university to apply it because it

saw an experiment in other universities in the same region and sector.

(P6) Hundred percent, observability matters. In our university we applied

many technologies with this. […] Say, this tool is helping you do your job

comfortably; you do it efficiently. If there is convincing evidence that it

was applied well in some area, then definitely, we will consider its use in

other areas. (P7)

Others were somewhat cautious about the effect of the observability factor

Ok, I can see, for example, KAUST or University of Taibah or any

university that has adopted the use blockchain technology. I may start to

see how their results and benefits are, but it’s not that it will be the

deciding factor or the only factor influencing our decision to adopt. (P2)

If we see the benefits of blockchain applications, it may not necessarily

be what we are looking for. We will first consider how blockchain is

applied and whether we use the same applications and systems, whether

it will benefit us the same. It’s a long process, actually, with many

variables involved. (P3)

Some interviewees were unsure about the overall effect:

I am not sure about the influence of observability. Maybe when the

decision-makers see that many universities in Saudi Arabia applied the

blockchain and had success, that may help them adapt to their university

because they know its benefits. Again, however, it is not clear right now.

(P4)

Adoption of new technologies is based more on politics. […] Most Saudi

universities are public, and they pursue a somewhat cautious

bureaucracy by waiting for some results and making observations of

some applications at other universities. Based on these experiences and

observations, technologies can be launched. But it will depend on the

bureaucratic procedures and state support. (P8)

Yet others denied the effect of observability altogether, arguing that their HEIs follow

their own technology adoption plans:

My university does not care whether other universities apply a new

technology or not. It is always searching for development on its own.

(P1) In my opinion, observing successful technology adoption may

motivate universities to apply it, but it is unnecessary, especially if a

university has a strong IT department and implements new technologies

regardless. (P5)

Therefore, the following themes emerged during the interviews with regard to

observability in relation to blockchain adoption by HEIs: 1) the effect of the

observability factor remains unclear; 2) observability will matter if blockchain benefits

can be observed both within and outside of a particular HEI; 3) it may take time for a

HEI to realise blockchain benefits by observing its effect; and 4) for some HEIs,

observability may not matter at all if they have a strong IT development strategy. Based

on the analysis, observability was retained in the final conceptual model, although its

effect was expected to be moderate.

Technology complexity was discussed by all interviewees in the context of blockchain

adoption by HEIs. The opinions, once again, were divided. Those who opined about the

influence of the complexity factor agreed that the impact of complexity would be

negative; that is, the more complex blockchain is considered to be, the weaker the

administrators’ desire to adopt it will be. However, disagreements arose about the

strength of this impact. Those who spoke of a strong impact argued that it would be

difficult for users to understand and use:

This factor will negatively impact technology adoption. When the

technology is challenging to use and understand and needs a specific

infrastructure and competencies, financially expensive, it may be a

reason for the delay in applying this technology. […] I see that this is the

case with blockchain at the moment. (P4)

I expect perceived complexity to be an important factor. Many institutions

still do not know exactly what difficulties they may face with this

technology and the weaknesses in this technology. (P5)

Complexity is a significant factor because blockchain is still emerging so

it could be hard for users to understand it. (P9)

Several interviewees also acknowledged the impact of complexity, although they

thought it would be moderate because of specialist and user training and the learning

process:

It is a possible factor, yes. Blockchain is ambiguous. However, I think the

impact of complexity can be negated, so it is a question whether there is a

specialist who can manage blockchain. (P2)

Blockchain needs effort in training academic staff, members, and

students. In the beginning, they may encounter problems, but with time

this will diminish. (P6)

There is some complexity, perceived complexity barrier in the beginning.

But it may not be as strong if users understand it quickly. So, it is a

matter of learning curve. (P10)

Yet others did not believe complexity would be a factor in the adoption process. They

argued that generally, in the age of technology, complexity is becoming archaic because

users are becoming increasingly technology savvy and because it is possible to hire

appropriate specialists at the institutional level:

Complexity as a factor? No, no, no. We live in the age of technology, and

complexity is a sentence that does not exist in our technology-driven life.

Users, especially young people, they are too well versed with all kinds of

emergent technologies. (P2)

No, complexity cannot stand in the way. Because the whole world is

almost electronic, it is impossible that the complexity of the blockchain

would be an obstacle to its adoption. Because we can use experts who

can arrange training for the staff you have or make the matter easier for

you. (P3)

Right now, I think the complexity of technology is not important. Because

each university has an IT department, the staff who work there are rather

familiar with new technology, and you can train them easily. We have no

problem. As long as these staff have been sufficiently trained, the

complexities, even if they exist, will not be an obstacle. (P8)

In the course of the analysis, the following themes emerged: 1) perceived blockchain

complexity negatively affects its adoption and it arises from the novelty of the

technology and users not being able to understand it outright; 2) the negative impact of

complexity can be diminished through training and learning; and 3) the stronger a HEI’s

IT department is believed to be, the lower the impact of complexity is expected to be.

The complexity factor was retained in the final conceptual model with the expected

negative impact on blockchain adoption.

The final technology factor discussed by the interviewees was the trialability of

blockchain. In general, the majority of study participants supported the effect of this

factor, although to different degrees. Those who argued in favour of a strong influence,

connected it to the necessity of trying any new potentially impactful technology on a

limited basis:

Of course, this factor is essential because any new technology must be

tested or experimented with to know if this technology is applicable and

to what extent and benefit. (P5)

Yes, it is necessary that technology can be tried out. For example, we

have three-phase trials in our university tech department for all new

technologies. If we see that it has good potential, it has a high chance of

being implemented. (P7)

Trialability is very, very, very important. Any new technology is always

applied on a limited scale after we see benefits from it and start applying

it in a broader form. (P8)

Those who believe that trialability only had a moderate degree of influence mostly

argued that almost any technology could be tried out, at least from the experience in

their HEIs:

Usually, the scenario that we have adopted uses any new technology that

we use in the infrastructure; whatever we need is a demo. We present it to

the decision-makers, whether the university director or the vice-dean in

the administration if they give us the approval and that it is complete. We

do not have any problem. (P2)

We do a trial for all applications. It is extensive and quick. But, as I said,

all applications are triable. (P3)

Trialability is a factor, to some extent. We, of course, test new

technologies to see what they help achieve and how. I do not, however,

recollect, any difficulties in trying and testing new technologies. (P6)

Finally, one interviewee did not believe that trialability should be a concern exactly

because new technologies are all tried routinely at her HEI:

No, trialability is an anachronism. We try all technologies that we

consider prospective, no exception. We test them quickly and then

transfer them to the other parties with the most specific features that

serve them.

(P10)

In the course of the analysis, the following themes emerged: 1) trialability is an

influential factor in blockchain adoption, although the strength of the influence is not

clear; 2) HEIs that routinely test emerging technologies for school applications are

better positioned to ignore the trialability factor.

All pattern codes and their underlying themes are summarised in Table 24, showing that

all five factors proposed within the Rogers (1995) model were retained for the final

analysis, although the expected strength of each factor varied.

Table 24: Data Summary for Technology Factors

Pattern Codes Themes Factor Decision

Relative

advantage -> blockchain

adoption -

Perceived as the strongest technology factor for

blockchain adoption;

catering to the growing amount of education

moving online; corresponding to specific

needs of HEIs in managing the process;

the advantage overall comes from being forward

looking, value adding technology with nextlevel

security mechanisms.

RETAINED in the

final model within

the technology

dimension. Strong

relationship to

adoption expected.

Compatibility ->

blockchain

adoption

compatibility is sought in all aspects of HEI

management: from services to administration

thereby making compatibility an important

factor in blockchain adoption; from a

technical standpoint, compatibility with the

existing systems may not be that important for

the decision makers.

RETAINED in the

final model within

the technology

dimension. Moderate

relationship to

adoption expected.

Observability ->

blockchain -

adoption

the effect of observability factor remains

unclear;

observability will matter if blockchain benefits

can be observed both within and outside of a

particular HEI; it may take time for a HEI to

realise blockchain benefits by observing its

effect;

for some HEIs, observability may not matter at

all if they have a strong IT development

strategy.

RETAINED in the

final model within

the technology

dimension. Moderate

relationship to

adoption expected.

Complexity ->

blockchain -

adoption

perceived blockchain complexity negatively

affects its adoption and it arises from the

technology novelty and users not being able

to understand it outright; the negative impact

of complexity can be diminished through

training and learning; the stronger a HEI’s IT

department is believed to be, the lower the

impact of complexity is expected to be.

RETAINED in the

final model within

the technology

dimension. Negative

relationship to

adoption expected.

Trialability ->

blockchain

adoption

trialability is an influential factor in blockchain

adoption, although the strength of the influence

is not clear;

HEIs that routinely test emerging technologies

for school applications are better positioned to

ignore the trialability factor.

RETAINED in the

final model within

the technology

dimension. Moderate

relationship to

adoption expected.

6.3.3.2 Organisational Factors

Data codes and the frequency with which organisational factors were mentioned are

listed in Table 25. The coding is ranked based on factor strength. It can be seen that top

management support was the most discussed and the most supported factor within the

organisational dimension. Organisational readiness was another factor with perceived

strong influence. However, organisational size was not as strongly perceived. Pattern

coding and theme development for each factor in relation to blockchain adoption based

on the interviews data are provided next.

Table 25: Coding Applied to Organisational Dimension Factors

Strength of Effect on Adoption

Code Meaning N Strong Moderate None Unclear

ORG-TMS Top Management

Support

10 9 1 - -

ORG-REA Organisational

Readiness

8 6 1 - 1

ORG-Size Organisational Size 10 3 - 4 3

Top management support was almost universally seen by the study participants as a

strong influential factor in blockchain adoption. As a matter of fact, it was considered

important in the context of all technologies. The interviewees discussed top

management support using the adjectives like “essential,” “necessary,” “significant” and

“most important.” The majority agreed that without top management support, it is

impossible to integrate new technologies in Saudi HEIs in general. Some typical

comments in this regard were:

Top management support is essential. We can't move and can't do

anything unless the senior management gives us authority and supports

us. (P1)

If a question about implementing some new technology arises, it is

necessary that it be with the support of the senior management. (P5) If

top management is not sufficiently convinced of new technologies, we

will have great difficulty implementing them. (P8)

For me, this is the most important factor. If there is no top management

support for a new technology, forget about it. (P9)

Some respondents also mentioned senior’s management general desire to try and adopt

new perspective technologies, although they must be convinced that those technologies

have good potential:

At my university, they support adopting new technology. However, you

have to convince them that this technology will serve them right. (P10)

We are working on digital transformation according to Vision 2030, so

top management is interested in new technologies. They will likely

consider blockchain, but you need to be convincing about it and its

specific advantages. (P7)

Everything that the top management supports makes things easier. If

there is support from the top management, I feel that it will be easy to

apply any new technology such as blockchain. But the question is how

you “sell” it to them. (P4)

Overall, in the course of the analysis, the following themes emerged: 1) top management

support is considered the strongest and most important organisational factor for

blockchain adoption; 2) top management support is generally regarded as necessary for

all types of technologies in Saudi HEIs; 3) top management in Saudi HEIs is generally

interested in new technologies, although they have to be given strong arguments about

the benefits of these technologies. The top management support factor was retained for

the final model in the analysis with an expected strong influence on blockchain

adoption.

Organisational readiness was also seen as a strong factor for blockchain adoption by the

study participants. The majority of the interviewees spoke about readiness in terms of

the human, technology, and financial resources necessary for new technology adoption.

This was usually supported by their experience in past and current technology projects

in their HEIs:

Based on the past and ongoing technology projects in our university, I

can say that financial resources, staff experience, and infrastructure

would all affect the adoption of blockchain technology. (P6)

New technologies other than blockchain have been applied at my

university. Educational institutions always have sufficient technological,

financial, human resources to apply new techniques. (P9)

For any new technology a university adopts, it is necessary to know the

particular resources that it needs. (P5)

Few respondents were less particular with regard to the role of organisational readiness

by mentioning that even without proper resources, institutions can still prepare for new

technology adoption by training and acquiring the resources in question:

Some universities may lack, for example, human resources. But then you

can develop a training plan and fill this gap. I think the same can be

applied to other resources as well. (P1)

The institution readiness factor can be remedied, solved, and dealt with

preparation, training, and the development of clear action plans for it by

equipping the institution. I mean, in terms of human or technical

resources. (P8)

In the course of the analysis, the following themes emerged in relation to the

organisational readiness factor: 1) organisational readiness is an important factor that

encompasses the human, financial, and technological resources necessary for technology

adoption; 2) institutions without a proper level of readiness can still adopt new

technologies by acquiring or developing such resources. Based on the analysis, the

organisational readiness factor was retained for the final analysis. A strong relationship

to blockchain adoption was expected.

Unlike the previously discussed organisational factors within the TOE framework,

organisational size was widely regarded as either a weak or insignificant factor by the

majority of interviewees. They argued that size will not matter if the institution has

proper management support and sufficient resources for new technology adoption:

The factor of the size of the educational institution does not make much

difference for as long as you have all the resources and management

support available. (P3)

I do not see it as very important. For example, now, we have banks and

municipalities that are using blockchain technology, or at least they are

trying it out. They did not care about the size of their institution. I do not

think it is a significant thing. I think that the most important thing for me

is the top management support and the institution's readiness. (P4)

My point of view is that the size of the institution is not essential because

if there is support and there is a ready environment, if there are human

staff ready if there is good material support, it does not matter whether

you are a 1,000-student institution or 10,000-student institution. (P9)

Other interviewees could not decide whether an institution of a particular size would

have an advantage in technology adoption:

Also, the factor of the size of the educational institution is also an

important factor, but it is not much because it is possible to apply the

technology to any sized institution. (P5)

Ok, size may matter, I do not know which one is better, however. I can see

positive aspects influencing new technology adoption by being either

small or large institution. (P8)

Granted, some interviewees still considered it a strong factor. However, there was no

consistency among them about which sized institution would be better positioned for

blockchain adoption:

The size of an organization affects the adoption of new technologies.

Smaller universities are more agile in adopting new technologies like

blockchain. (P1)

If the university has many students, I will try to adopt new technology

like the blockchain because my goal is to speed up and improve

efficiency of certain administrative processes. Accordingly, larger

universities will adapt new technologies like blockchain faster. (P2)

I wish it would equally apply to small and big institutions. However, it

will benefit larger institutions to a larger degree. Therefore, they will

pursue it more willingly. (P6)

In the course of the data analysis, the following themes emerged in relation to

organisational size and its potential impact on blockchain adoption: 1) organisational

size is generally considered as a weak, non-significant factor; 2) there are advantages for

both small and large HEIs in adopting new technologies like blockchain; 3) there is no

consistency about which organisational size is better positioned for new technology

adoption. After careful analysis of the relevant data, it was decided to remove

organisational size from the organisational dimension of the final model. However,

given its possible, although unclear influence on blockchain adoption, the organisational

size variable will be considered among the control variables in the quantitative data

analysis.

All pattern codes and their underlying themes relevant to the organisational dimension

factors are summarised in Table 26. It can be seen that two of three factors proposed in

the original TOE framework were retained for the final analysis. For both top

management support and organisational readiness, a strong relationship to blockchain

adoption was expected.

Table 26: Data Summary for Organisational Factors

Pattern Codes Themes Factor Decision

Top Management

Support ->

blockchain

adoption

perceived as the strongest organisational

factor for blockchain adoption;

generally regarded as necessary for the

adoption of all types of technologies in

Saudi HEI; top management in Saudi

HEIs are generally interested in new

technologies, although they have to be

given strong arguments about the

benefits of these technologies.

RETAINED in the

final model within

the organisational

dimension. Strong

relationship to

adoption expected.

Organisational Readiness ->

blockchain -

adoption

organisational readiness is an important

factor that encompasses human,

financial, and technological resources

necessary for technology adoption;

institutions without a proper level of

readiness can still adopt new

technologies by acquiring or

developing such resources.

RETAINED in the

final model within

the organisational

dimension. Strong

relationship to

adoption expected.

Organisational Size ->

blockchain adoption

organisational size is generally

considered as weak, non-significant

factor; there are advantages for both

small and large HEIs in adopting

new technologies like blockchain;

there is no consistency about which

organisational size is better positioned for

new technology adoption.

REMOVED from

the final model’s

organisational

dimension. Will be

considered among

the control

variables.

6.3.3.3 Environmental Factors

Data codes and the frequency with which environmental factors are mentioned are listed

in Table 27. The coding is ranked based on factor strength. It can be seen that nearly all

interviewees discussed the effect of all the considered environmental factors on

blockchain adoption. Existing regulations were considered as the most influential

factors in this regard, although the strength of such influence was not always clear.

Government support was also seen as an important factor, although some respondents

disagreed. Finally, the effect of the peer pressure factor was rather ambiguous with an

even distribution of supporters and opponents on its effect. Pattern coding and theme

development for each factor in relation to blockchain adoption based on the interview

data are provided next.

Table 27: Coding Applied to Environmental Dimension Factors

Strength of Effect on Adoption

Code Meaning N Strong Moderate None Unclear

ENV-REG Existing regulations 10 5 3 - 2

ENV-GVT Government support 10 4 4 2 -

ENV-PPR Peer pressure 8 3 2 3 -

The potential effect of existing regulations was discussed by all interviewees in relation

to blockchain adoption. The majority of interviewees agreed that the effect was present,

although opinions differed regarding its strength. The proponents of a strong effect

argued that new technologies must meet the existing requirements and that this is

especially true for public HEIs in Saudi Arabia. Another argument was that HEIs have

internal regulations that must be considered as well:

True, government regulations are barriers to the fast implementation of

modern technologies. This is especially related to new technologies. (P4)

Obviously, in Saudi Arabia we have the Ministry regulating technology

matters. We need to carefully consider such regulations, and blockchain

will not be an exception in this case. (P3)

The regulations definitely represent one of the major barriers. Sometimes

it can be a real hindrance to new technology adoption. However, at the

very least, we need to comply with the regulations existing within the

university. (P8)

Some interviewees, however, argued that the regulations apply to specific technologies

only and that it is still possible to try them out even though on a limited basis:

With regard to existing regulations, I should say you need to look at what

specific technologies are regulated and how. Maybe they are regulated,

maybe they are not, maybe only partially regulated. So, it depends, really.

(P5)

Whether regulations represent a really significant barrier is

questionable. I am sure, and we do this at our school, you can still try

new technologies on a limited basis, no problem. Regulations, they play a

role when you want to launch a new technology on a wide scale,

officially. (P6)

Whether regulations are important, I’d say yes, but probably not as much

as you may think. We, at least, have not had many issues with regulations

with such things as the cloud, AI, or even IoT. We look at new tech and

what laws apply to it, but it is usually not that strict. (P7)

Finally, some respondents could not determine whether the effect would be present in

relation to blockchain:

As you know, blockchain is a new technology. It is very new, but it has

been already tried in other fields like finance, for example. I think

regulations exist in relation to cryptocurrency, but I am not sure whether

there are any for blockchain in education. So, I cannot speak of this

confidently. (P9)

Regulations do matter. Speaking of their effect on blockchain right now,

maybe both yes and no. Yes, because I know that blockchain is somehow

regulated on the financial side. No because I haven’t heard of some

restrictions for our field. And, besides, you know, a couple of universities

already use this technology in Saudi Arabia. I do not know how it is

regulated. (P10)

Based on the analysis of the qualitative data, the following themes emerged in relation

to the effect of existing regulations on blockchain adoption: 1) regulations likely matter,

and they have a negative effect on technology adoption; 2) both external and internally

developed regulations will likely matter; 3) the effect of regulations may be diminished

because they may apply to specific technologies and their uses and because HEIs may

still experiment with technologies on a limited basis. It was decided to retain the

existing regulations factor in the final model. The expected effect on blockchain

adoption would be moderately negative.

The effect of government support was discussed by all interviewees as well. The

majority of study participants argued in favour of the positive effect of government

support, although to differentextents. At the same time, there were respondents who did

not consider government support a significant factor in blockchain adoption. The

proponents of a strong effect argued that government support started to matter more

during the pandemic as the HEIs were in dire need of new technologies to offer high

quality education services online. Further, they argued that government support means

more resources and less red tape in the adoption process:

Government support means a lot in new technology adoption. They

provide the resources, sometimes financial, sometimes advice. This is

very true for public universities in Saudi Arabia. But I also think it plays

an increasing role for private schools as well. (P1)

Cooperation from the Ministry of Education with educational institutions

in applying technologies and keeping pace with the times is essential.

During the COVID-19 pandemic, the Ministry played a major role in the

transformation of distance education. I think this will continue. A lot of

trust has been built between the Ministry and the universities in the past

few years. (P9)

If there was one positive effect of COVID, I would say it was the

increasing role of government support. With more and more educational

services moving online, government support matters more as well. We

have noticed this with the application of new technologies. (P10)

Those who felt that government support was a moderate factor proposed that HEIs often

rely on internal sources in developing and applying new technologies, although they

would welcome any kind of support coming from the government.

Speaking of government support, I’d say it is always welcome. However,

we have been mostly implementing new technologies ourselves. (P5) I

believe the government can be supportive in new technology

implementation and adoption. At the same time, I think internal

resources are more important. We rely on them, and this would likely be

the case with blockchain. (P6)

Finally, those who did not think of government support as an influential factor in

blockchain adoption argued that government agencies are generally slower and more

cautious when it comes to new technology adoption. Conservative policies and

bureaucracy would prolong the wait for adequate support in terms of resources:

Speaking of anything new, government agencies are rarely supportive

outright, at least in our field. New technologies are not always seen as a

top priority. If we asked for additional technology or, for example, we

adopted new technology, it can be presented to them. However, it will

require time in terms of providing us money and taking time until

approval begins. (P10)

The Ministry of Education is not an obstacle when it comes to new

technology adoption, but they are not very supportive either. This is how

I see it. There is too much old, conservative thinking, too much

bureaucracy involved. They react slowly to innovations, we need to rely

on ourselves. (P3)

Based on the analysis of the qualitative data, the following themes emerged in relation

to the effect of government support on blockchain adoption: 1) government support is

seen as a positive thing, although it is not always expected; 2) government support is

more likely at later stages of innovation when the benefits of new technologies become

obvious; 3) conservative policies and bureaucracy are seen as the main elements in the

process. It was decided to retain the government support factor in the final model. The

expected effect on blockchain adoption would be moderately positive.

The role of the peer pressure factor in blockchain adoption caused the most

disagreement in the opinions of the interviewees. Those who argued about a strong

effect of peer pressure talked about competitiveness and a strong motivation when

adopting new technologies like blockchain:

Peer pressure is an essential factor because we are working in the same

sector, so we need to remain competitive. If other universities apply a

new technology, this is a competitive advantage. We must respond

quickly. (P6) Peer pressure is a significant factor. If blockchain is

adopted by us of some other university to a great positive effect and we

find it really useful and helpful, then for sure it matters because the goal,

in general, is to improve development for the better. (P9)

Yes, our university looks all the time if other universities use some new

tools, looks at the results, how they would have benefited from it if they

had found great benefit from it, apply it here as well. (P10)

Other study participants acknowledged the effect of peer pressure but thought it did not

play a very important role. The interviewees pointed to peer pressure as a reason to

make decisions about riskier types of technologies with unclear potential:

I think, peer pressure is a catalyst, although to some extent only. I mean,

it can stimulate the adoption of certain technologies, but mostly those

where benefits are not clear. In the case of other technologies, where the

effects are known and visible, you do not wait for others, you explore

them timely. (P4)

Peer pressure helps convince the decision makers if some risky

technology is under review. Say you do not have any significant problems

and you do not intend to replace the existing technologies right now. Why

should I take risks? But then I see that others are trying it, ok I will play

a waitand-see game. If they are successful, it will pressure the decision-

makers to agree to use this technology. However, it is not that we see

others implementing it and go for it right away. (P2)

An equal proportion of the respondents did not believe peer pressure had any substantial

effect. They based their opinions on HEIs’ internal programs of technology

development:

Ours is a competitive field, for sure. To remain competitive, we cannot

afford the wait-and-see games. (P1)

As for peer pressure, no, it’s not a factor, because ultimately every

institution has its own program and a plan for technologies. (P5) With

regard to peer pressure, it is unnecessary, because every university has

its development plans, and they know what their technical capabilities

are and know what the unique technical needs are in it. New

technologies, therefore, are directed at filling the internal service and

administration gaps, not to meet technology capacity of others. (P8)

Based on the analysis of the qualitative data, the following themes emerged in relation

to the effect of peer pressure on blockchain adoption: 1) peer pressure could be a factor

in HEIs using technology as a point of competition; 2) peer pressure probably has a

stronger impact in relation to new, untried technologies with few observational benefits;

3) for tried, proven technologies, peer pressure is unnecessary because their adoption is

implemented within HEIs’ internal development plans. It was decided to retain the peer

pressure factor in the final model. The expected effect on blockchain adoption would be

moderately positive, although it could differ based on to what extent the respondents

perceive blockchain to be a risky technology.

All pattern codes and their underlying themes relevant to the environment dimension

factors are summarised in Table 28. In the end, all factors proposed in the original TOE

framework were retained for the final analysis. A strong negative relationship was

expected between the existing regulations and blockchain adoption and moderately

positive relationships were expected for the relationships between government support

and blockchain adoption as well as peer pressure and blockchain adoption.

Table 28: Data Summary for the Environment Factors

Pattern Codes Themes Factor Decision

Existing

Regulations ->

blockchain adoption

regulations likely matter, and they have a

negative effect on technology adoption; both

external and internally developed regulations

will likely matter; the effect of regulations

may be diminished because they may apply

to specific technologies and their uses and

because HEIs may still experiment with

technologies on a limited basis.

RETAINED in the

final model within

the environment

dimension. Strong

negative relationship

to adoption expected.

Government

Support ->

blockchain adoption

government support is seen as a positive

thing, although it is not always expected;

government support is more likely at later

stages of innovation when the benefits of

new technologies become obvious;

conservative policies and bureaucracy are

seen as the main elements in the process.

RETAINED in the

final model within

the environment

dimension. Moderate

positive relationship

to adoption expected.

Peer Pressure >

blockchain

adoption

peer pressure could be a factor for HEIs

using technology as a point of competition;

peer pressure probably has a stronger impact

in relation to new, untried technologies with

few observational benefits;

for tried, proven technologies, peer pressure

is unnecessary because their adoption is

implemented within HEIs’ internal

development plans.

RETAINED in the

final model within

the environment

dimension. Moderate

positive relationship

to adoption expected.

6.3.3.4 Quality Factors

Three quality-related factors were discussed by the study participants: expected

improvements to education quality, expected improvements to administration processes,

and a reduction in graduate unemployment levels. Data codes and the frequency with

which quality factors are mentioned are listed in Table 29. The coding is ranked based

on the discussed factor strength. It can be observed that all three factors were perceived

to have a strong influence on blockchain adoption by the interviewees. However, while

the study participants clearly identified the role of blockchain in improving employment

opportunities for graduates, they rarely separated education service quality from school

administration quality, as will be shown later. Pattern coding and theme development for

each factor in relation to blockchain adoption based on the interviews data are provided

next.

Table 29: Codes Applied to Quality Dimension Factors

Strength of Effect on Adoption

Code Meaning N Strong Moderate None Unclear

QUAL-EMP Reduced

unemployment

10 7 1 - 2

QUAL-EDU Quality of education

services

8 7 1 - -

QUAL-ADM Quality of school

administration

8 7 1 - -

The majority of the interviewees agreed on the strong effect of blockchain’s potential to

reduce unemployment which, in turn, may prompt its faster adoption. The inherent

features of blockchain, such as immutability and transparency, were commonly

discussed as important factors in this regard:

Blockchain in my opinion is more transparent and more credible

regarding employment. There will be a direct match of skills based on

blockchain-backed credentials. Indeed, I see that a reduction of student

unemployment is one of the critical quality factors in blockchain

adoption. (P4)

Blockchain technology will significantly improve education quality

because blockchain education credentials will be based on impeccable

evidence. Ledger-verified credentials cannot be altered or forged. (P8)

Accordingly, the interviewees pointed to the ease and speed of the job application

process and the verification of student credentials using blockchain certificates:

In the increasingly digitalised world, students can apply online for many

jobs around the world, but this process needs the verification of skills.

Traditional certificates are not that easy to confirm, it requires time and

effort. Blockchain-based credentials will make the process so much

easier. This is also another strong reason why universities should

consider blockchain. (P5)

When there is a need, the search for employees is rapidly using

blockchain technology because their data and achievements are saved,

accessible, and easily verified, so hiring is faster. (P9)

One interviewee also pointed to an increased motivation for students to excel in their

academic endeavours because blockchain certificates record learning progress that

cannot be manipulated or altered:

Sure, blockchain improves the performance and operational objectives

and reduces student unemployment. It helps in reducing the

unemployment level by motivating students to develop and show their

experiences and authenticate their experiences and certificates. That will

encourage them. Determining the work shall be according to the

competence and experience of this student. There is no manipulation in

it. (P6)

Some interviewees, however, could not determine the exact way that blockchain could

reduce unemployment for college graduates. In all cases, they pointed to the advantages

for the early adopters with an unclear effect for the majority of students:

Maybe there is a connection to employability, maybe it will give an

advantage in the labour market. However, that will depend on

blockchain’s use. I am not sure many students will be eager to try it. (P2)

Blockchain may increase employment chances for people who master

and understand technology. But what about those who do not support

this technology, do not want to develop themselves, are satisfied with

their approximate situation, satisfied with their situation and so on? I do

not see the automatic benefits for all these people, so only for those who

are blockchain savvy. And maybe universities will understand it. (P3)

In the course of data analysis, the following themes emerged with regard to the

relationship between improvements in student employment and blockchain adoption: 1)

inherent, unique features of blockchain may assist students in the employment market;

2) the most likely advantages provided by blockchain in the labour market are ease of

the applicant’s review process and credibility of education achievements; 3) blockchain

records may additionally motivate students to excel in studies hence becoming better

prepared for professional life; and 4) there is some doubt whether blockchain will

benefit all students, not only those who embrace the technology early. Based on the

analysis, reduced unemployment was retained in the final model with a moderate

positive effect on blockchain adoption expected.

The interviewees also expressed views on the strong positive effect of blockchain on

education quality and, hence, on blockchain adoption by HEIs. However, with the

exception of 1 respondent, all of them talked about education quality improvements and

administration improvements in the same terms and context:

I expect strong contribution of blockchain to quality of education. I do

not mean courses only, I mean the entire process of managing,

administering, delivering services. (P3)

Blockchain technology will have a major role in improving the quality of

education. […] during the Covid-19 pandemic, distance education has

become dominant. The more universities become reliant on distance

education, the stronger role blockchain may play in improving its quality.

(P4)

Sure, blockchain will likely improve the quality of education. Since the

goal is always to improve development for the better, the application of

blockchain technology will be the best solution in administrative work

and services because it is more transparent and more credible and will

help in the process of archiving and disposing of digital documents,

reducing the likelihood of falsifying and losing papers. . (P9)

Yes, we can say that blockchain will improve the quality of education

because the whole education system will be almost online and provide

online resources and lectures to students at any time and from any place

that the student can deliver to, such as micro credentials, how it is

possible for the student to enrol in short courses and courses from any

university around the world and to take them online. (P5)

Based on the analysis of the qualitative data, the following themes emerged in relation

to the effect of improvements in education quality on blockchain adoption: 1) quality of

education is seen as a composite of service quality and administration quality; 2) the

quest for improving the quality of education will prompt HEIs to adopt blockchain as

HEIs become more digitalised. Based on the analysis, it was decided to retain the

combined quality of education services and quality of education administration into a

single factor in the final model. The new factor is titled improving education quality.

The expected effect on blockchain adoption would be positive.

All pattern codes and their underlying themes relevant to the quality dimension factors

are summarised in Table 30. For the final model, three original factors were collapsed

into two with the separate effects expected for reduced unemployment and improved

education quality.

Table 30: Data Summary for Quality Factors

Pattern Codes Themes Factor Decision

Reducing

graduates’

unemployment -> -

blockchain adoption

inherent, unique features of blockchain

may assist students in labour market; the

most likely advantages provided by

blockchain in the labour market are ease of

the applicant’s review process and

credibility of education achievements;

blockchain records may additionally

motivate students to excel in studies hence

becoming better prepared for professional

life;

there is some doubt whether blockchain will

benefit all students, not only those who

embrace the technology early.

RETAINED in the

final model within

the quality

dimension. Strong

positive relationship

to blockchain

adoption expected.

Improving quality

of services -> -

blockchain

adoption

Improving HEI

administration ->

blockchain adoption

quality of education is seen as a composite

of service quality and administration

quality; the quest for improving the

quality of education will prompt HEIs to

adopt blockchain as HEIs become more

digitalised.

CONVERGED into

a single improving

education quality

factor. Strong

positive relationship

to blockchain

adoption is expected.

6.3.3.5 Barriers to Adoption

The interviewees discussed various barriers to the adoption of blockchain technology.

With a total of 11 barriers mentioned, it was possible to divide them into three groups:

individual, organisational, and external. Table 31 shows the data coding for these

barriers within these three groups with most frequently mentioned barriers first. Despite

the multitude of barriers, only some of them were thoroughly discussed by a sufficient

number of interviewees. Moreover, some barriers were similar to or paralleled the

previously mentioned factors, all of which allowed them to be converged into a single

factor. Therefore, in the final model, only four barriers were retained. Pattern coding and

theme development for each factor in relation to blockchain adoption based on the

interview data are provided next.

Table 31: Codes applied to Blockchain Adoption Barriers

Strength of Effect on Adoption

Code Meaning N Strong Moderate Unclear

Ind

INDB-LKN

ividual Barriers

Lack of knowledge about

technology

8 2 1 5

INDB-PSC Privacy and security

concerns

7 5 1 1

INDB-LNG Language concerns 7 5 - 2

INDB-RSK Risk avoidance 2 - 1 1

Organisational Barriers ORGB-

LSP Lack of specialists 5 3 2 -

ORGB-INF Lack of appropriate

infrastructure

3 2 1 -

ORGB-LFN Lack of finances 2 1 1 -

ORGB-LRD Lack of general readiness 1 1 - -

External Barriers

EXT-FINA Association with finance

only

8 6 1 1

EXT-LVIS Lack of visibility 4 2 - 2

In total, 4 individual barriers to blockchain adoption were discussed by the interviewees.

A lack of knowledge about blockchain was the most commonly mentioned factor (8

times), although the study participants failed to mention the strength of its effect.

Specifically, many interviewees agreed that blockchain is still an emergent technology,

especially in Saudi Arabia, and therefore, it may take time before HEIs will start to

adopt it:

A possible barrier is, of course, that there is insufficient information

about blockchain technology. This lack of knowledge prevents

administrators from exploring it. (P5)

Maybe insufficient information about the blockchain is one of the

barriers. Like I said, many haven't heard of it. It may be a mistake or a

shortcoming on our part here in the Arab world. I hear very little about

blockchain at the industry conferences and seminars. This is the strong

reason why blockchain adoption may take time. (P7)

It is still an emerging technology that is still mysterious and has not been

widely applied, so this is the challenge for universities. (P9)

One of the main things preventing blockchain adoption in colleges and

universities is ignorance about the technology and its advantages. The

level of ignorance is especially high here, in Saudi Arabia. (P10)

Based on the analysis of the interviews, the following themes regarding the relationship

between a lack of knowledge about blockchain and its adoption by Saudi HEIs emerged:

1) lack of knowledge generally prevents technology adoption; 2) blockchain is currently

poorly understood and rarely applied; 3) however, it is not clear to what extent this may

prevent its adoption. Due to the high frequency with which it was mentioned, it was

decided to include lack of knowledge as a barrier to blockchain adoption in the final

model. The expected direction of the effect would be negative, although the strength of

the effect is difficult to predict.

Privacy and security concerns related to blockchain adoption were expressed by 7

interviewees. There was agreement that this could be a strong factor preventing

blockchain adoption by some HEIs, despite the blockchain features that actually

enhance both. This is based on the perception of all new technologies which are not well

understood yet:

The concerns of privacy and security are always strong with new

technologies, blockchain will not be an exception. People who know the

technology will know that it is secure. However, it is difficult for those

who have little idea how blockchain works. And many decision makers in

Saudi Arabia’s colleges and universities are like that, unfortunately. (P2)

Privacy and security concerns are very important, but they are based on

misunderstanding of technologies. Look, we had this problem with cloud

applications before. There was a suggestion for using cloud computing,

and they said ‘no, it does not work because of security’ OK guys, security

in the cloud is greater than the existing data center. It took them several

years to understand this. The same will be the case with blockchain. (P3)

Privacy and security are strong in blockchain, but it is possible to

imagine people having fears because they do not understand the

technology […] this may be considered a serious obstacle to adopting

this technology. (P5)

Based on the analysis of the interviews, the following themes regarding the relationship

between privacy and security concerns about blockchain and its adoption by Saudi HEI

emerged: 1) privacy and security concerns are typical for new technologies among

Saudi school administrators; 2) it may take years to realise the real privacy and security

potential similar to cloud computing; 3) privacy and security concerns cause serious

reservations regarding blockchain adoption. Due to the high frequency with which it

was mentioned, it was decided to include privacy and security concerns as a barrier to

blockchain adoption in the final model. The expected direction of the effect would be

strongly negative.

Language concerns emerged as a surprising but strong factor that could prevent

blockchain adoption in Saudi HEIs. The interviewees argued that there is very little

information about blockchain in Arabic. They also expressed concerns that it would be

difficult to find non-foreigners to implement blockchain if needed:

Language could be a serious problem. If we want to learn about new

technology, how it works, what its benefits are, we would rather read

about it in our native language, which is Arabic. However, there is so

little information about blockchain in Arabic! (P1)

A very novel technology blockchain might be, but where do we get

specialists to implement and run it? I doubt you can find many good

blockchain specialists in Saudi Arabia. There are no courses in

blockchain, there is no technical literature in Arabic. Unless you get

some specialists with good English skills, there is little information about

it. (P3) Also, please, consider language issues. I would say that 95%

maybe of all information on blockchain is in English. Most of staff here

do not speak or read English. (P4)

Sure, you can find a lot of information on blockchain on the internet. You

can even go through some online courses. The problem? It’s all in

English or other languages, not in Arabic. And this is a problem because

not many people are capable of reading lest understanding some difficult

technical parts of it in English. (P7)

It is often the case with innovative technologies coming to Saudi Arabia

from abroad. Most systems are native English languages, the client of

universities slowly adopts it if he does not support Arabic. There is very

little information about blockchain, this is a very huge drawback for

adoption. (P10)

Based on the analysis of the interviews, the following themes regarding the relationship

between language concerns and blockchain adoption by Saudi HEI emerged: 1) there is

little information about blockchain in Arabic; 2) administrators would feel

uncomfortable implementing technologies without thorough descriptions available in

their native language; 3) the language barrier in relation to blockchain may be a reason

for the shortage of good blockchain specialists. Due to high frequency with which it was

mentioned, it was decided to include language concerns as a barrier to blockchain

adoption in the final model. The expected direction of the effect would be strongly

negative.

Fewer interviewees (4 in total) discussed general risk avoidance by school

administrators as a potential impediment to blockchain adoption. While there were

opinions that risk avoidance could be a strong factor preventing the adoption of new

technologies, those opinions largely matched the ones expressed during the discussions

of the peer pressure factor. Specifically, the interviewees believed that risk avoidance

with new technologies is one of the major reasons that school administrators may wait

and see the effects of technology on other HEIs:

School administrators here in Saudi Arabia are very risk averse. They

would rather miss opportunities than take the blame for something they

implement and which does not work properly. Therefore, they would

generally prefer to be slow with new technologies like blockchain and

just wait before they become widespread. (P5)

I believe that risk avoidance by university tops could be a barrier.

Basically, they would rather watch how the untried technologies work if

they are implemented by others. They would rather not take risks on their

own. (P8)

Because risk avoidance was very close semantically to discussions on other factors and

because it was discussed by a few interviewees only, it was decided to not include this

factor in the final model among the potential barriers to blockchain adoption by Saudi

HEIs.

Four types of organisational barriers to blockchain adoption were identified during the

interviews: lack of specialists, lack of appropriate infrastructure, lack of finances, and

lack of general readiness of a HEI. None of these barriers, however, was distinctively

different from the factors discussed in the organisational dimension. Specifically, all

these factors were reverse of the organisational readiness aspects:

The existing infrastructure could be a challenge. I mean, the current

systems may not be well suited for blockchain. (P2)

University infrastructure may not be commensurate with blockchain

technology. It could make universities unable to use them. (P4)

Technical resources, finance resources, staff experiences, you name it.

This subject needs further study, what is required for its successful

implementation? What kind of resources? (P6)

As I’ve already mentioned, technical readiness is important. If there is

insufficient technical competence of staff, especially since we are talking

about new technology like blockchain, then this will be an issue. (P8)

Even if blockchain is really good and useful, if your organisation is not

generally ready for new technology, then you would have to pass on it, at

least for some time. (P7)

One of the reasons for not adopting new technology could be the lack of

cadres. But let’s say, you have technically competent staff. Maybe you do

not have sufficient budget for implementation or maybe the university

will not assign sufficient funds for some reason. So, human and financial

resources are the key: you do not have either one in sufficient amount, no

successful blockchain adoption will be possible. (P10)

Because there were no distinctive aspects of the organisational level barriers from the

requirements mentioned in the organisational factors dimension, it was decided that

these barriers would not be included in the final model.

The final group of barriers to adoption discussed by the interviewees were external

factors, of which HEIs have little or no control. Blockchain’s association with finance

only and its lack of visibility were the two barriers mentioned. Association with finance

emerged as a unique, technology-specific barrier for blockchain adoption discussed by 8

study participants. Almost all of them considered blockchain’s association with finance

as a strong barrier to adoption. Specifically, the respondents talked about blockchain

being associated with cryptocurrencies and its perceived absence of useful applications

for education:

For many people, even those who know about blockchain, this is

something from the investing field. Bitcoin, cryptocurrency, maybe

financial transactions. But few really think of it as an application for the

educational field. (P1)

True, one of the main challenges is that we often talk about blockchain in

the context of bitcoin. And, actually, the majority of information about

blockchain is about this. I understand this is because of the incredible

growth of bitcoin in the past years. Because of this, who would think of

blockchain for education instead of gaining personal wealth? (P2)

I would also argue that […] blockchain is strongly linked to the financial

field in human minds. This is what it was created for in the first place.

Sure, it is moving to other fields as well, but the majority still believe it is

a financial instrument, not an instrument for the educational field. This,

in turn, is a barrier to adoption, a very strong barrier. I mean, until

human perceptions of blockchain change. (P6)

You know, some universities in Saudi Arabia already use blockchain. But

you go and ask people how blockchain is used, and almost everyone will

answer that it is a bitcoin instrument. They will find it hard to speak of

blockchain applications for education. (P10)

Based on the analysis of the interviews, the following themes regarding the relationship

between blockchain’s association with finance and its adoption by Saudi HEIs emerged:

1) blockchain is strongly associated with finance; 2) many people, even those familiar

with blockchain, find it difficult to talk about its useful applications or benefits for the

educational field; and 3) this type of thinking prevents forward-looking actions directed

at blockchain adoption by HEIs. Due to high frequency with which it was mentioned, it

was decided to include association with finance as a barrier to blockchain adoption in

the final model. The expected direction of the effect would be strongly negative.

Fewer respondents talked about the lack of blockchain’s visibility, and only 2

acknowledged that this could be a serious barrier:

How do you make a technology adopted? Through knowledge of course.

You see it, see its potential, you adopt it. With blockchain technology, all

the world still has ignorance about it. It is not somewhere on social

media or in the news, I mean for general audiences. It is for a limited

number of specialists for now. (P1)

There is no blockchain visibility, really. Compared to cloud computing,

for example. Almost everyone knows what it is. But, again, a few years

ago, it was the same situation. Today, since the cloud is everywhere,

everyone knows about it, everyone adopts it increasingly. For

blockchain, no such visibility is a problem. (P6)

The other two interviewees took a more cautious approach in this regard by noting that

HEIs may adopt blockchain not for the masses, but to serve their own needs thereby

reducing the need for visibility:

I admit that blockchain has not yet received due prominence, especially

in the financial field. Few know about it. But again, if we talk about a

university, like our university, for example, we adopt technologies to

serve internal needs, be it services or administration, or something else.

So, the mass visibility of technology is not a big barrier. (P8)

I am not sure that blockchain should be really a widespread, much

discussed technology before schools start using it. The reason is simple:

blockchain applications may be used to solve internal technology gaps.

In some aspects, yes, it must be visible, especially when it involves users

such as students. In other aspects, however, like improving security

aspects of school servers, this may not be required. (P9)

Because only a few interviewees discussed the lack of blockchain visibility as a barrier

and because there was no consensus about it being a significant barrier, this factor was

not included in the final model for the study.

Table 32 summarises the pattern codes, themes, and decisions regarding the inclusion of

the potential barriers to blockchain adoption in the final model. From the original 10

barriers, 4 were retained for the analysis. The decisions were based on the number of

interviewees discussing the factors, the depth of the discussions, and the uniqueness of

each factor to improve the descriptive power of the model.

Table 32: Pattern Codes, Themes, and Decision Regarding Barriers to Adoption

Factors

Pattern Codes Themes Factor Decision

Lack of

knowledge ->

blockchain

adoption

lack of knowledge generally prevents technology adoption;

blockchain is currently poorly understood and rarely applied;

however, it is not clear to what extent this may prevent its

adoption.

INCLUDED in

the final model.

Privacy and

security concerns -

> blockchain

adoption

privacy and security concerns are typical for new

technologies among Saudi school administrators; it

may take years to realise the real privacy and security

potential similar to cloud computing; privacy and

security concerns cause serious reservations regarding

blockchain adoption.

INCLUDED in

the final model.

Language

concerns ->

blockchain

there is little information about blockchain in Arabic;

administrators would feel uncomfortable implementing

technologies without thorough descriptions available

INCLUDED in

the final model.

adoption in their native language; the language barrier in

relation to blockchain may be a reason for the shortage

of good blockchain specialists.

Risk avoidance ->

blockchain

adoption

school administrators are generally risk averse when it

comes to new technologies;

they would prefer to wait and see how their peers do with

the technology in question.

EXCLUDED

from the final

model

Lack of specialists -

> blockchain

adoption

being a new technology for education, blockchain

specialists in this field are scarce; it may take time and

resources to achieve the required level of technical

competence with the existing staff.

EXCLUDED

from the final

model

Lack of appropriate

infrastructure ->

blockchain

adoption

if a HEI lacks the appropriate infrastructure, new technology

has little chance of being implemented; appropriate

infrastructure must include systems and processes

conducive to blockchain use which is often not the case.

EXCLUDED

from the final

model

Lack of finances >

blockchain

adoption

the lack of funds prevents investments in new

technologies and processes; school administrators may not

issue funds if they are not fully confident that the

technology will succeed.

EXCLUDED

from the final

model

Lack of general

readiness ->

blockchain

adoption

general readiness arises from a combination of human,

financial, and infrastructure factors; if these factors,

wholly or partially, are not present, there will be issues

with technology adoption.

EXCLUDED

from the final

model

Association with

finance only ->

blockchain

adoption

blockchain is strongly associated with finance; many

people, even those familiar with blockchain, find it

difficult to talk about its useful applications or benefits for

educational field; this type of thinking prevents forward-

looking actions directed at blockchain adoption by HEIs.

INCLUDED in

the final model.

Lack of visibility >

blockchain

adoption

blockchain lacks visibility unlike other technologies such

as cloud computing;

visibility is apparent especially important in the

education field; visibility may matter less if a HEI seeks a

technology to improve its internal processes.

EXCLUDED

from the final

model

6.4 Finalised Conceptual Model

Based on the analysis of qualitative data, the original conceptual model was amended to

take into account the insights acquired from the Saudi HEI specialists. In sum, the

following changes were applied:

1) The organisational factors dimension was reduced to two factors by

eliminating the size variable which was considered either weak or

nonsubstantial in blockchain adoption by the interviewees;

2) The quality dimension was reduced from three to two factors by combining

improvements to education and improvements to administration into a single

factor titled improvements to education quality;

3) The human factors dimension was dissolved with the innovativeness feature

included in the top management support factor and technical competence

included into organisational readiness factor;

4) Four out of ten factors were retained in the barriers to adoption dimension.

Other factors were eliminated or integrated with other factors based on the

interviewees’ input.

The final conceptual model which served as a basis for quantitative data analysis is

presented in Figure 25. The model consists of five dimensions: technology factors,

organisational factors, environment factors, quality factors, and barriers to adoption. The

influence of sixteen variables across these dimensions on blockchain adoption in Saudi

HEIs will be analysed based on quantitative data from a large-scale survey.

Figure 25: Finalised Conceptual Model for the Study

6.5 Finalised Questionnaire

Based on the finalised conceptual model presented in Figure 25, the instrument for the

quantitative data analysis was developed. Previously validated TOE, DOI, and quality

assurance questionnaires were taken as a basis for item development in this study.

Because of the lack of research on blockchain adoption in higher education, the

literature pertaining to the adoption of other new technologies (cloud computing, smart

campus etc.) in higher education was consulted. Table 33 lists the study constructs with

the corresponding studies from which the items were extracted.

Table 33. List of Variables and Sources for Item Development

Context Variable Sources

Technological

Relative advantage

Compatibility

Complexity

Crosby et al. (2016), Iansity and Lakhani

(2017), Lustenberger et al. (2021)

Guo and Liang (2016), Shrier et al.

(2016) Drescher (2017), Kouhizadeh et

al. (2021), Wong et al. (2020)

Observability Lou and Li (2017), Lustenberger et al.

(2021), Rauchs et al. (2019)

Trialability Clohessy and Acton (2019), Lustenberger

et al. (2021), Schmitt et al. (2019)

Technology Factors

Relative advantage

Compatibility

Observability

Complexity

Trialability

Organisational

Factors

Top Management

Support

Organisational

Readiness

Quality Factors

Improvements to

Education Quality

Reduction of

Graduates’

Unemployment

Environmental

Factors

Regulations

Government Support

Peer Pressure

Barriers to

Adoption

Lack of Knowledge

Association with

Finance Only

Privacy and Security

Language Concerns

Blockchain

Adoption in

Saudi HEIs

Organisational

Top management support

Institution readiness

Duan et al. (2020), Kouhizadeh et al.

(2020), Wang et al. (2020),

Mendling (2017), Kouhizadeh et al.

(2021), Webster and Gardner (2019)

Institution size Crosby et al. (2016), Iansity and Lakhani

(2017), Morabito (2017)

Environmental

Existing regulations

Peer pressure

Hiran and Henten (2019); Mendling et al.

(20180, Wong et al. (2019)

Clohessy and Acton (2019), Kouhizadeh

et al. (2021)

Government support Shrier et al. (2016), Tapscott and Tapscott

(2016)

Quality

Service quality improvement

Unemployment reduction

Al-Ramahi and Odeh (2020); Harvey

(2007)

Al-Ramahi and Odeh (2020), Lindman et

al. (2017)

Barriers

Lack of knowledge

Privacy and security concerns

AlTaei et al. (2019), Delaghani et al.

(2022); Li et al. (2019)

Kokina et al. (2017), Lindman et al.

(2017), Reddick et al. (2019)

Association with finance Ma and Fang (2020); Raimundo and

Rosario (2021)

Language barriers Alenezi et al. (2021), Hiran (2021)

Adoption Intent to adopt blockchain

Clohessy and Acton (2019), Delghani et

al. (2022), Lustenberger et al. (2021),

Webster and Gardner (2019)

The original questionnaire is presented in Appendix D. The questionnaire was divided

into the following parts:

•Part 1: A cover letter which includes a consent or approval document and

information regarding the research, information on the researcher conducting the

study, the researcher’s supervisors and their contact information. It also stated that

the UTS Research Ethics and Integrity Policy has been followed in all stages of the

research.

•Part 2: General information to describe the sample participants and the institutions

they represented, such as gender, age, position held in their institution, and their

level of familiarity with blockchain technology. The following institution data were

collected: size (based on the number of students), type (private or public), current

technologies in use (from the list), and the current level of blockchain adoption.

•Part 3: Technology factors of blockchain adoption. This section included five

factors identified within the DOI theory: relative advantage (5 items), compatibility

(3 items), trialability (3 items), observability (4 items) and complexity (4 items).

•Part 4: Organisational factors of blockchain adoption. This section included two

factors identified in the TOE theory: top management support (4 items) and

institutional readiness (3 items).

•Part 5: Environmental factors of blockchain adoption. This section included three

factors identified in the TEO theory: existing regulations (3 items), government

support (2 items) and peer pressure (3 items);

•Part 6: Quality factors of blockchain adoption. This section included two factors

identified in the relevant research on quality in higher education: education quality

improvements (3 items) and student employment improvements (4 items).

•Part 7: Barriers to blockchain adoption. This section included four factors identified

in the DOI and TOE frameworks: lack of knowledge (4 items), association with

finance (4 items), privacy and security concerns (4 items) and language concerns (2

items).

•Part 8: Adoption of blockchain, which was represented by 2 items.

6.6 Chapter Summary

This chapter presented the results of the first phase of the research which involved

interviews with the decision makers holding administrative and IT positions in Saudi

colleges and universities. The total sample included 10 individuals, with the limit

reached at the point of qualitative data saturation (no meaningful new insights emerged

from 11th interview). The interview participants discussed the current state of blockchain

adoption in education and the factors which they believed were important for the

blockchain adoption process in Saudi HEIs. The data analysis followed the Miles and

Huberman (2019) methodology and involved qualitative data reduction, theme

development, and pattern matching. In general, the respondents viewed blockchain

applications in higher education positively, although they also noted its limited use at

the moment. Diplomas and certificates emerged as the key area of blockchain

applications in higher education. Therefore, the finalised questionnaire was adjusted

specifically to this application of blockchain technology.

The second part of the analysis focused on refining the original research framework. The

interviewees discussed their views regarding the factors included in the model and

proposed additional factors that they believed were relevant. The analysis led to the

integration of some factors into others, eliminating some redundant constructs and

adding two additional factors of language concerns and association of blockchain with

finance only. The refined model included 14 factors instead of 21 in the original

framework. The refined framework served as the basis for quantitative instrument

development and an analysis based on a large-scale survey. The next chapter presents

the survey results.

7 Research PHASE2: Model Evaluation

7.1 Introduction

This chapter presents the analysis of the collected survey data. Section 7.2 provides the

descriptive analysis of the data, the survey response rate and sample structure. Section

7.3 provides the preliminary data analyses to ensure good data quality and to check for

possible issues with bias, normality, validity and reliability. Section 7.4 offers the results

of the structure equation modelling (SEM) with the analyses covering model validation

and hypotheses testing. Section 7.5 summarises the results within the formulated

conceptual framework.

7.2 Descriptive Analysis

In total, 504 online survey questionnaires were submitted from the target population. To

ensure good data quality, a set of exclusion criteria was applied. Specifically, the

following submissions were excluded from the analysis:

1. Incomplete surveys;

2. Surveys submitted in a very short time period (a minimum reasonable time to

read, comprehend, and answer the questionnaire items was set at 6 minutes as

determined during the pilot test of the questionnaire);

3. Surveys containing the same responses to all items (reverse items were included

in the survey to ensure that the respondents had read all the items and answered

thoughtfully).

After the exclusion criteria were applied, 289 completed questionnaires were retained

for analysis. Given the estimated population size of 2,000 individuals, the margin of

error was 5.3% (Dattalo, 2008).

7.2.1 Sample Characteristics

The descriptive statistics of the sample includes basic information about the educational

institutions where the respondents work, their positions in those institutions, and their

level of blockchain knowledge.

Regarding the type of higher education institution (HEI), 260 respondents (90%) were

employed in public colleges and universities and 29 respondents (10%) in private HEIs.

The majority of respondents (n = 133, 46.0%) were from large size HEIs of 15,000

students or more, followed by midsize HEIs of 5,000-14,999 students (n = 87, 30.1%)

and smaller HEIs with fewer than 5,000 students (n = 69, 23.9%). The results are

presented in Table 34.

Table 34: Organisational Characteristics of the Sample

Institution by Type of Funding

n %

Public 260 90

Private 29 10

Total 289

Institution by Size

100

n %

15,000+ students 133 46.0

5,000 - 14,999 students 87 30.1

< 5,000 students 69 23.9

Total 289 100

In terms of the individual characteristics of the respondents, the majority (n = 112,

38.8%) were senior IT personnel and CTOs, followed by senior administrative

personnel such as presidents, vice-presidents, deans and the members of the board (n =

91, 31.5%). The remaining respondents (n = 86, 29.8%) represented mid-level IT and

administrative positions. Of the respondents, the vast majority were familiar with

blockchain technology, having either good (n = 125, 43.3%) or some knowledge (n =

108, 37.4%) about it. Only a small number of respondents (n = 56, 19.4%) indicated

having little to no knowledge about blockchain. The results are presented in Table 35.

Table 35: Individual Characteristics of the Sample

Respondents by Position

n %

Senior IT 112 38.8

Senior Administrative 91 31.5

Other 86 29.8

Total 289

Knowledge of Blockchain

100

n %

Good 125 43.3

Some 108 37.4

Little or none 56 19.4

Total 289 100

7.3 Preliminary Analysis

Provided below is the preliminary analysis of the collected data. The goal of the

preliminary analysis is to ensure data integrity prior to running the inferential statistics

tests. Accordingly, the preliminary analysis included: screening for missing data, general

statistics of the scale items, tests of data normality assumption, outlier screening, and bias

tests. Further, the data were checked for validity, internal consistency, and

multicollinearity.

7.3.1 Missing Data Analysis and Scale Items’ Statistics

The first step in the analysis was to check for the missing items and report the means

and the standard deviations of the scale items. Appendix E shows the SPSS results of the

descriptive statistics analysis for all 61 scale items. It can be seen that all items were

represented by 289 responses. Given that the total number of respondents was 289, there

was no indication of missing data.

The means and standard deviations of all scale items are also reported in Appendix E. It

can be observed that all response means were above 3, which represents the middle

value of the 7-point Likert-scale. Therefore, on average, the respondents treated all

items positively. A measure of standard error of the sample mean (Altman & Bland,

2005) was applied for all items to test for the degree of variability. The values ranged

between 6.8% and 11.2%, indicating relatively low levels of variation in the data.

7.3.2 Data Normality

Normality refers to the tendency of the collected data to match the normal distribution

(Hair, Black, Babin, & Anderson, 2018; Johnson & Wichern, 2007). Testing for

normality determines whether the data should be analysed with parametric or non-

parametric tests (Kline, 2015). Whereas large samples are typically believed to

demonstrate the normal distribution shape in general, certain items may still exhibit

deviations due to outliers. The normality tests for the collected data in this study were

performed by measuring the skewness and kurtosis levels of the scale items. Several

guidelines exist regarding the admissible values of skewness and kurtosis. Some authors

suggest that both skewness and kurtosis should be within +2 value to perform normal

univariate distribution tests (George & Mallery, 2010). Others propose looser acceptable

values for kurtosis at +7 (Byrne,

2010; Hair, Black, Babin, & Anderson, 2018). Yet others argued that due to the

robustness of structural equation modelling tests, data should be considered normal for

the purpose of analysis if its skewness is within +3 range and kurtosis is within +10

range (Brown, 2006; Kline, 2015).

The normality tests performed for the scale items used in this research are reported in

Appendix E. The results demonstrated a skewness range between -1.338 and 0.442 and

kurtosis range between -1.148 and 2.454. As such, the results were very close to the

most stringent normality ranges suggested in the literature, as discussed above.

Therefore, the collected data were deemed normal for the purpose of this research and

analysis.

7.3.3 Outlier Screening

Outliers are data points that lie way outside of the main data pattern (Osborne &

Overbay, 2004). Outliers can be natural: that is, a small percent of unusually different

observations are expected in a large population. However, in some cases, outliers can

result from measurement or data collection errors. Both types of outliers may distort the

data and affect the statistical analyses results; however, non-natural outliers are more

dangerous since they do not represent real observations. For this reason, this study

screened for outliers in the collected dataset. The screening was performed using the z-

score technique where scale items are transformed into standardised scores. The

obtained coefficients show how many standard deviations a datapoint is above or below

the mean. A rule of thumb is that outliers lie above the absolute value of a standardised

score of 3.29 which cuts off 0.1% of all data points (Martin & Bridgmon, 2012;

Tabachnik & Fidell, 2018).

Appendix F shows the z-scores for all scale items in the dataset. It can be seen that

several variables possessed data points that could be considered outliers. However, the

maximum number of outliers for a single item was 7 (QF_QI_1), which represented

only 2.4% of the total observations. According to Tabachnik and Fidell (2018),

removing outliers is recommended only when there is strong evidence that they are

beyond what could normally be observed in a sample data. The presence of 2.4% or

fewer responses containing unusual data can be interpreted as normal. Since the study

used a 7-point Likert scale for answers, it is not unusual that a few respondents took

either very positive or very negative perspectives on most of the issues. In fact, this

could be evident through the consistent presence of some respondents across most of the

outlier scores. Specifically, respondents 2, 3, 5, 98 and 252 consistently provided very

negative responses to relative advantage, trialability and quality of service whereas the

vast majority provided positive responses. Therefore, to preserve the integrity of the

data, it was decided to keep all responses, including the outliers, for the data analysis.

7.3.4 Non-Response Bias Tests

With 289 valid submissions out of 504, the non-response rate was about 42.7%. This

raised concerns about possible non-response bias. According to Coderre et al. (2004),

non-response bias represents the potential differences between the population

representatives who respond to surveys and those who do not. Accordingly, if such

differences exist, a generalisation of the survey results onto the population becomes

problematic (Werner, Praxedes, & Kim, 2007). A common technique to check for

nonresponse bias is described by Armstrong and Overton (1977) where partial

respondents and later survey respondents are treated as proxies to non-participants and

their responses are compared to the early survey participants. A series of t-tests were

conducted to check for potential differences in scale item responses between 30 early

survey participants and 30 late survey participants. Because no statistically significant

differences were observed, it was concluded that the collected data were not prone to

non-response bias.

7.3.5 Common Method Bias Tests

Whereas the non-response bias test deals with potential participation issues, the common

method bias test analyses possible data distortions due to data collection methods. In the

literature, common method bias is defined as variance stemming from the procedures

rather than the actual variables that the research measures should represent (Lowry &

Gaskin, 2014; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). In other words, the

choice of methods may inflate the strength of the actual relationships between the

variables in a study. To examine the collected data for the common method bias,

Harman’s test was used with an unrotated factor analysis for all variable items and a

cutoff point of 50% for a single factor (Podsakoff et al., 2003). The SPSS analysis

produced 11 factors with the largest one accounting for 32.457% variance (Appendix

G), well below the established cut-off point. Therefore, it was concluded that common

method bias was not an issue with the collected survey data.

7.3.6 Factor Analyses

Factor analysis is an important step in research because it validates the instrument used

for quantitative data collection and determines how well the survey items align with the

formulated constructs. The two commonly used methods to do this are exploratory

factor analysis (EFA) and confirmatory factor analysis (CFA) (Kline, 2015). EFA seeks

to determine the data structure and a maximum amount of variance in the factors

whereas CFA seeks to validate the existing models to ensure they fit with the existing

study context (Fabrigar & Wegener, 2012; Widaman, 2012). As such, EFA is considered

more appropriate in the early research stages or for testing the newly developed

instruments whereas CFA is better suited for previously confirmed measures and

models.

While survey data collection in this study relied mostly on the previously developed

scales, the instrument itself was translated into Arabic which arguably makes it “new”.

Moreover, the model in this study includes several items in the quality of education

construct which have not been applied before. For this reason, both EFA and CFA were

used. There are also recommendations in the literature in support of this approach. To

enhance the quality of factor analysis, some authors recommend applying EFA for

model specification and later apply CFA to cross-validate the model (Cabrera-Nguyen,

2010; Taherdoost, Sahibuddin, & Jalaliyoon, 2014; Worthington & Whittaker, 2006).

This study followed those recommendations.

Factor analyses were performed for five multi-variable dimensions in the conceptual

model: technology factors (5 variables), organisational factors (2 variables),

environmental factors (3 variables), quality factors (2 variables), and barriers to

adoption (4 variables). The EFA were performed using principal axis factoring

extraction with a promax rotation which assumes that the items are correlated

(Tabachnik & Fidell, 2018). The Kayser-Meyer-Olkin (KMO) test was performed to

check the data fit for factor analysis aiming for values 0.7 and above (Kline, 2015). For

individual items, communality scores and cross-loadings were used to determine their

inclusion in the final model. Acceptable communality scores were considered above 0.4

(Osborne, Costello, & Kellow, 2008). Acceptable cross-loadings were considered in the

range below/above +0.2 for the items loading on more than one factor (Ding & Shen,

2017). Following Nunnally (1978) and Hair et al. (2018), individual item loadings onto

their corresponding factor should be ≥0.5 while the average of all item loadings onto

their corresponding factor should be ≥0.7. Finally, the internal consistency of the data

for each construct was measured with Cronbach’s alpha and item-rest correlation with

acceptable values ≥0.7 and ≥0.3 respectively (Nunnally, 1978).

A CFA was performed after the EFA for each dimension in the model. According to

Kline (2015), for models exceeding 200 cases, the chi-square values could be

misleading. Therefore, the model fit was estimated based on the number of additional

parameters such as comparative fit index (CFI, ≥0.9), normality fit index (NFI, ≥0.9),

Tucker-Lewis index (TLI, ≥0.9), root mean square error of approximation (RMSEA,

≤0.1), and standardized root mean square residual (SRMR, ≤0.1) (Hair, Black, Babin, &

Anderson, 2018; Kline, 2015; Williams, Hartman, & Cavazotte, 2010). The tests for

convergent and discriminant validity were performed using average variance extracted

(AVE). For convergent validity, the acceptable AVE score was ≥0.5 whereas for

discriminant validity, the square root AVE scores had to be below the constructs’ cross-

correlations (Fornell & Larcker, 1981; Hair, Black, Babin, & Anderson, 2018).

7.3.6.1 Technology Dimension

The technology dimension was represented by five factors: relative advantage,

compatibility, trialability, observability, and complexity. The results of the EFA are

shown in Table 36. The rotation converged in 6 iterations onto 5 constructs with an

acceptable KMO = 0.891 and cumulative 78.376% variance. All communality scores

were above the established 0.4 threshold. Two potentially problematic items were

observed with cross-loadings onto two factors: TF_COM_1 and TF_TRI_3. However,

only TF_COM_1 showed a cross-loading within the +0.2 range which made it a

candidate for elimination. Moreover, its loading power on both factors was relatively

weak (.451 and .432 respectively). Therefore, it was decided to drop TF_COM_1 item

from further analysis to avoid discriminant validity issues. The EFA analysis without the

dropped item showed no possible issues with factor loadings or relative parameters

(Table 37).

Table 36: Results of the First EFA for Technology Dimension Variables

Component

Items Communality

Score 1 2 3 4 5

TF_RA_1 .753 .890

TF_RA_2 .830 .935

TF_RA_3 .757 .897

TF_RA_4 .747 .765

TF_RA_5 .721 .743

TF_COM_1 .672 .451 .432

Pattern Matrix a

TF_COM_2 .847 .857

TF_COM_3 .844 .954

TF_COM_4 .675 .687

TF_TRI_1 .733 .745

TF_TRI_2 .841 .864

TF_TRI_3 .745 .411 .674

TF_OB_1 .876 .950

TF_OB_2 .884 .995

TF_OB_3 .748 .837

TF_OB_4 .659 .679

TF_COX_1 .748 .881

TF_COX_2 .838 .902

TF_COX_3 .881 .927

TF_COX_4 .876 .910

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization. KMO = 0.891, p < .01 a.

Rotation converged in 6 iterations. Suppressed values below 0.4

Table 37: Results of Final EFA for Technology Dimension Constructs

Component

Items Communality

Score 1 2 3 4 5

TF_RA_1 .756 .826

TF_RA_2 .827 .883

TF_RA_3 .765 .878

TF_RA_4 .770 .778

TF_RA_5 .746 .756

TF_COM_2 .839 .820

TF_COM_3 .870 .940

TF_COM_4 .644 .796

TF_TRI_1 .812 .887

TF_TRI_2 .838 .884

TF_TRI_3 .712 .764

Pattern Matrix a

TF_OB_1 .863 .932

TF_OB_2 .882 .987

TF_OB_3 .756 .840

TF_OB_4 .681 .703

TF_COX_1 .748 .882

TF_COX_2 .839 .902

TF_COX_3 .880 .928

TF_COX_4 .877 .911

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization. KMO = 0.884, p < .01 a.

Rotation converged in 6 iterations. Suppressed values below 0.4

The maintained data were analysed for internal consistency. All five constructs

demonstrated high reliability scores for both individual items (CITC) and construct

variables (Cronbach’s alpha). Therefore, all items and constructs were retained for the

CFA. Table 38 provides a summary of the reliability analyses for all retained technology

dimension items and constructs.

Table 38: Reliability Analysis of the Technology Dimension Items

Item Wording Mean St.D CITC Cronbach’s

alpha

Relative Advantage

TF_RA_1 Blockchain will speed up the university certificate process. 5.60 1.426 .756

.917

TF_RA_2 Blockchain will make the certificate issue process easier. 5.74 1.338 .836

TF_RA_3 Blockchain will reduce the costs of certificate issue. 5.70 1.260 .782

TF_RA_4 Blockchain will make certificates more secure. 5.93 1.226 .790

TF_RA_5 Blockchain will streamline certificate verification. 5.90 1.160 .770

Compatibility

TF_COM_2 Blockchain is suitable for issuing university certificates. 5.52 1.259 .769

.861 TF_COM_3 Blockchain is suitable for university certificate verification. 5.64 1.199 .796

TF_COM_4 Blockchain is compatible with certificate issue processes. 5.29 1.327 .656

Trialability

TF_TRI_1 It will be possible to set a trial for blockchain-based certificate

issue

5.47 1.264 .679 .841

TF_TRI_2 Out school would prefer to trial and test blockchain before full

scale implementation.

5.51 1.267 .794

TF_TRI_3 A successful trial for blockchain-based certificate issue and

verification will be key to its implementation.

5.81 1.162 .652

Observability

TF_OB_1 There is sufficient information about blockchain applications 4.19 1.815 .848 .909

for education.

TF_OB_2 There is sufficient information about blockchain application for

education certificates.

4.33 1.804 .858

TF_OB_3 We are aware of blockchain use for university certificates. 4.46 1.746 .778

TF_OB_4 Blockchain has so far demonstrated benefits for education

institutions.

Complexity

4.97 1.481 .708

TF_COX_1 Blockchain is conceptually difficult to understand. 3.86 1.695 .737

.927

TF_COX_2 Blockchain benefits for school certificates are hard to exploit. 3.61 1.663 .840

TF_COX_3 Blockchain technology is difficult to use for school certificates. 3.41 1.654 .879

TF_COX_4 It is difficult to master blockchain applications for school

certificates.

3.38 1.688 .869

The results of the CFA for the retained variables and items are presented in Table 39.

The model showed a good fit (CFI=0.939, NFI=0.911, TLI=0.927, RMSEA=0.082,

SRMR=0.079). All standardised regression coefficients were above 0.7 which represents

strong loadings. The convergent and discriminant validity of the data were estimated

based on the AVE scores. The AVE was estimated at 0.727, above the 0.5 threshold.

Therefore, convergent validity was established. Further, the AVE square root estimate of

0.853 was above the highest correlation of the variables (r = 0.800, relative advantage –

compatibility). Therefore, discriminant validity was also established.

Table 39: CFA for Technology Dimension Constructs and Items

Item Standardised Highest

Regression AVE √𝐀𝐕𝐄 Construct Model Fit

Weights Correlation

TF_RA_1 .827

TF_RA_2

TF_RA_3

TF_RA_4

TF_RA_5

TF_COM_2

TF_COM_3

TF_COM_4

TF_TRI_1

TF_TRI_2

TF_TRI_3

TF_OB_1

.829

.830

.854

.734

.759

.881

.873

.789

.871

.752

.900 0.727 0.853 0.800

CFI = 0.939

TLI = 0.927

NFI = 0.911

RMSEA = 0.085//

SRMR = 0.079

TF_OB_2 .930

TF_OB_3 .774

TF_OB_4 .737

TF_COX_1 .937

TF_COX_2 .953

TF_COX_3 .849

TF_COX_4 .786

7.3.6.2 Organisation Dimension

The organisation dimension was represented by two factors: top management support

(TMS) and institution readiness (IR). The EFA results are presented in Table 40. The

rotation converged in 3 iterations onto 2 constructs with an acceptable KMO = .914 and

cumulative 78.272% variance. All communality scores were above the established 0.4

threshold. Two potentially problematic items were observed: OF_IR_1 and OF_IR_5

both of which cross-loaded significantly onto two factors. Because the cross-loading

difference for both items fell into the +0.2 range, these items were dropped from further

analysis to avoid discriminant validity issues. The final EFA run without the deleted

items is shown in Table 41. No issues with loadings or relevant parameters were present.

Table 40: Results of First EFA for Organisation Dimension

Pattern Matrixa

Items Communality Score Component

1 2

OF_TMS_1 .700 .793

OF_TMS_2 .873 .889

OF_TMS_3 .760 .750

OF_TMS_4 .857 .840

OF_IR_1 .653 .489

OF_IR_2 .790 .875

OF_IR_3 .836 .804

OF_IR_4 .826 .803

OF_IR_5 .746 .493 .512

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization, KMO = .914, p < .001 a.

Rotation converged in 3 iterations. Suppressed values below 0.4

Table 41: Results of the Final EFA for Organisation Dimension Variables

Pattern Matrixa

Component

Items Communality Score

1 2

OF_TMS_1 .714 .845

OF_TMS_2 .887 .941

OF_TMS_3 .774 .875

OF_TMS_4 .852 .922

OF_IR_2 .805 .895

OF_IR_3 .850 .919

OF_IR_4 .823 .905

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization, KMO = .873, p < .001 a.

Rotation converged in 3 iterations. Suppressed values below 0.4

The remaining items and constructs were analysed for internal consistency. Both top

management support and institutional readiness constructs demonstrated high reliability

scores for both individual items (CITC) and construct variables (Cronbach’s alpha).

Therefore, all remaining items and constructs were retained for the CFA. Table 42

provides a summary of the reliability analyses for all retained organisation dimension

items and constructs.

Table 42: Reliability Analysis for Organisation Dimension Constructs and Items

Item Wording Mean St.D CITC Cronbach’s

alpha

Top Management Support

OF_TMS_1 Our top management provides timely and sufficient

information about new technology implementation in

our institution.

5.02 1.610 .724

OF_TMS_2

OF_TMS_3

Our top management provides strong leadership when it

comes to technology adoption.

Our top management is likely to consider the adoption

of blockchain for university certificates as strategically

important.

5.34

5.23

1.420

1.404

.881

.784

.915

OF_TMS_4 Our top management supports new technology

implementation for the institution.

Institutional Readiness

5.52 1.326 .853

OF_IR_2 Our institution has sufficient financial resources to

integrate blockchain for certificates.

5.31 1.425 .750

OF_IR_3 Our institution has sufficient technical capacity for

blockchain adoption.

5.22 1.546 .831 .895

OF_IR_4 Our institution has sufficient human resource capacity

to implement blockchain technology for certificates.

5.11 1.595 .801

The results of the CFA for the retained variables and items are presented in Table 43.

The model showed a good fit (CFI=0.939, NFI=0.911, TLI=0.927, RMSEA=0.082,

Regression Weights

.938

.852

.897

.755

.835

.919

.821

SRMR=0.079). All standardised regression coefficients were above 0.7 which represents

strong loadings. The convergent and discriminant validity of the data were estimated

based on the AVE scores. The AVE was estimated at 0.679, above the 0.5 threshold.

Therefore, convergent validity was established. Further, the AVE square root estimate of

0.824 was above the correlation of the variables (r = 0.751). Therefore, discriminant

validity was also established.

Table 43: CFA for Organisation Dimension Constructs and Items

Item Standardised Construct

AVE √𝐀𝐕𝐄 Correlation Model Fit

OF_TMS_1

OF_TMS_2

CFI = 0.983

OF_TMS_3 TLI = 0.973

OF_TMS_4 0.679 0.824 0.751 NFI = 0.975

OF_IR_2 RMSEA = 0.086

SRMR = 0.080

OF_IR_3 OF_IR_4

7.3.6.3 Environmental Factors Dimension

The external factors dimension was represented by three factors: existing regulations

(ER), government support (GS), and peer pressure (PP). The EFA results are presented

in Table 44. The rotation converged in 5 iterations onto 3 constructs with an acceptable

KMO = .824 and cumulative 84.846% variance. All communality scores were above the

established 0.4 threshold. No cross-loading issues were observed for the items.

Therefore, all items and constructs were analysed for internal consistency.

All three constructs in the dimension demonstrated high reliability scores for both

individual items (CITC) and construct variables (Cronbach’s alpha). Therefore, all items

and constructs were retained for the CFA. Table 45 provides a summary of the reliability

analyses for all retained external factors dimension items and constructs.

Table 44: EFA for the Environmental Factors Dimension

Pattern Matrixa

Component

Items Communality Scores

1 2 3

EF_ERS_1 .880 .965

EF_ERS_2 .915 .968

EF_ERS_3 .734 .791

EF_GS_1 .828 .798

EF_GS_2 .880 .972

EF_PP_1 .843 .897

EF_PP_2 .864 .936

EF_PP_3 .843 .908

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization. KMO = .824, p < .001 a.

Rotation converged in 5 iterations. Suppressed values below 0.4

Table 45: Reliability Analysis for the Environmental Factors Dimension Constructs and Items

The results of the CFA for the retained variables and items are presented in Table 46.

The model showed a good fit (CFI=0.972, NFI=0.961, TLI=0.960, RMSEA=0.089,

SRMR=0.081). All standardised regression coefficients were above 0.7 which represents

strong loadings. The convergent and discriminant validity of the data were estimated

based on the AVE scores. The AVE was estimated at 0.823, above the 0.5 threshold.

Therefore, convergent validity was established. Further, the AVE square root estimate of

Item Wording Mean St.D CITC Cronbach’s

alpha

Existing Regulations

EF_ERS_1 The Ministry of Education dictates to us what

technologies to adopt.

4.71 1.654 .838

EF_ERS_2 The Ministry of Education dictates to us what

technologies to use.

4.71 1.631 .886 .906

EF_ERS_3 Blockchain is tightly regulated for organizations in Saudi

Arabia.

Government Support

4.48 1.646 .719

EF_GS_1

EF_GS_2

Our institution receives subsidies for new technology

implementation.

The Saudi government promotes new technology

applications in education institutions like ours. Peer

Pressure

5.03

5.49

1.408

1.315

.661

.796

EF_PP_1 We are willing to use blockchain for certificates if we see

that other institutions use it.

5.43 1.400 .819

EF_PP_2 We will consider implementing blockchain for certificates

if it becomes popular.

5.45 1.359 .833

.911

EF_PP_3 We will consider implementing blockchain for certificates

if we see that other institutions benefit from doing so.

5.54 1.304 .817

0.907 was above the highest correlation of the variables (r = 0.633, existing regulations

– government support). Therefore, discriminant validity was also established.

Table 46: CFA for the Environmental Factors Dimension Constructs and Items

CFI=0.972

NFI=0.961

TLI=0.960

0.823 0.907 .633

RMSEA=0.089

SRMR=0.081

7.3.6.4 Quality Factors

The quality factors dimension was represented by two factors: education quality

improvements (QI) and employment improvements (EI). The EFA results are presented

in Table 47. The rotation converged in 3 iterations onto 2 constructs with an acceptable

KMO = .879 and cumulative 83.246% variance. All communality scores were above the

established 0.4 threshold. No cross-loading effects were observed for the items;

therefore, all items and constructs were retained.

Table 47: EFA for the Quality Factors Dimension

Pattern Matrixa

Items Communality Score Component

1 2

QF_QI_1 .870 .928

QF_QI_2 .911 1.035

Item Standardised

Regression Weights AVE √𝐀𝐕

𝐄

Highest Construct

Correlation Model Fit

EF_ERS_1 .793

EF_ERS_2 .974

EF_ERS_3 .900

EF_GS_1 .810

EF_GS_2 .816

EF_PP_1 .884

EF_PP_2 .892

EF_PP_3 .863

QF_QI_3 .769 .708

QF_QE_1 .735 .618

QF_QE_2 .868 1.025

QF_QE_3 .902 .970

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization, KMO = .879, p < .001 a.

Rotation converged in 3 iterations. Suppressed values below 0.4

Both education quality improvements and employment improvements constructs

demonstrated high reliability scores for both individual items (CITC) and construct

variables (Cronbach’s alpha). Therefore, all remaining items and constructs were

retained for the CFA. Table 48 provides a summary of the reliability analyses for all

retained organisation dimension items and constructs.

Table 48: Reliability Analysis for Constructs and Items in the Quality Factors

Dimension

Item Wording

Education Quality Improvements

Mean St.D CITC Cronbach’s

alpha

QF_QI_1 Blockchain use for certificates will improve the

quality of education.

5.57 1.342 .836

QF_QI_2 Blockchain use will improve the quality of school

administration.

5.69 1.258 .864 .908

QF_QI_3 Blockchain use for certificates aligns well with our

school’s mission and goals.

Employment Improvements

5.65 1.255 .754

QF_QE_1

QF_QE_2

Our school considers new technologies as a way to

reduce student unemployment after graduation.

Blockchain technology has the potential to reduce

student unemployment.

5.26

5.20

1.438

1.440

.812

.885

.917

QF_QE_3 Blockchain technology will streamline the

employment process for students after graduation.

5.41 1.372 .791

The results of the CFA for the retained matrices initially demonstrated a poor model fit

(CFI=0.962, NFI=0.934, TLI=0.956, RMSEA=0.139, SRMR=0.111). A further analysis

of the standardised residual covariances matrix showed that QF_QE_3 item was

problematic and a potential source of construct validity. After removing this item from

the model, the CFA improved to an acceptable level (CFI=0.984, NFI=0.967,

TLI=0.980, RMSEA=0.089, SRMR=0.091). All standardised regression coefficients

were above 0.7 which represents strong loadings. The convergent and discriminant

Regression Weights

.947

validity of the data were estimated based on the AVE scores. The AVE was estimated at

0.796, above the 0.5 threshold. Therefore, the convergent validity was established.

Further, the AVE square root estimate of 0.892 was above the correlation of the

variables (r = 0.733). Therefore, discriminant validity was also established. The results

of the analysis are shown in Table 49.

Table 49: CFA for the Quality Dimension Constructs and Items

Item Standardised Highest Construct

AVE √𝐀𝐕𝐄 Correlation Model Fit

QF_QI_1

QF_QI_2

QF_QI_3

QF_EI_1

QF_EI_2

QF_EI_3

.892

.954

.874

.786

.828

.871 .796 .892 .733

CFI=0.984

NFI=0.967

TLI=0.980

RMSEA=0.089

SRMR=0.091

7.3.6.5 Barriers to Adoption

The barriers to adoption dimension was represented by four factors: lack of knowledge,

association with finance, privacy and security concerns, and language concerns. The

results of the EFA are shown in Table 50. The rotation converged in 5 iterations onto 4

constructs with an acceptable KMO = 0.811 and cumulative 84.500% variance. All

communality scores were above the established 0.4 threshold. No cross-loadings for the

items were observed. Therefore, all items and constructs were retained for further

analysis.

Table 50: EFA for the Barriers to Adoption Dimension

Pattern Matrixa

Component

Items Communality

Score 2 3 4

BF_LOK_1 .911

BF_LOK_2 .927 .965

BF_LOK_3 .922 .947

BF_LOK_4 .888 .947

BF_AWF_1 .666 .803

BF_AWF_2 .801 .828

BF_AWF_3 .748 .858

BF_AWF_4 .672 .729

BF_PAS_1 .880 .938

BF_PAS_2 .914 .964

BF_PAS_3 .834 .898

BF_PAS_4 .827 .903

BF_LC_1 .921 .942

BF_LC_2 .918 .932

Extraction Method: Principal Component Analysis.

Rotation Method: Promax with Kaiser Normalization. KMO =

0.811, p < .01

a. Rotation converged in 5 iterations. Suppressed values below 0.4

The maintained data were analysed for internal consistency. All four constructs

demonstrated high reliability scores for both individual items (CITC) and construct

variables (Cronbach’s alpha). Therefore, all items and constructs were retained for the

CFA. Table 51 provides a summary of the reliability analyses for all barriers to adoption

dimension items and constructs.

Table 51: Reliability Analysis for the Barriers to Adoption Constructs and Items

Item Wording Mean St.D CITC Cronbach’s

alpha

Lack of Knowledge

BF_LOK_1 We still know little about blockchain applications for

university certificates. 5.13 1.450 .918

.968

BF_LOK_2 We still know little about the technical implementation of

blockchain for university certificates. 5.13 1.472 .933

BF_LOK_3 We still know little about the success of blockchain for

university certificates. 5.10 1.497 .927

BF_LOK_4 We still know little about how to integrate blockchain

with the certificate issue/verification process. 5.10 1.493 .898

Association with Finance

BF_AWF_1 Blockchain is primarily a financial instrument. 4.79 1.559 .552

.819

BF_AWF_2 Blockchain so far has been used in finance only. 4.25 1.740 .733

BF_AWF_3 Blockchain is designed for financial transactions

primarily. 4.83 1.444 .679

BF_AWF_4 There has been little application of blockchain beyond

the financial industry. 4.38 1.673 .615

Privacy and Security Concerns

BF_PAS_1 We do not feel secure about sharing institution data on a

blockchain platform. 3.50 1.803 .894

.948

BF_PAS_2 We do not feel secure about sharing student data on a

blockchain platform. 3.57 1.809 .923

BF_PAS_3 We are not sure about the policies and regulations

regarding data on a blockchain platform. 3.89 1.911 .845

BF_PAS_4 Student data on blockchain is a potential privacy

concern. 3.83 1.896 .841

Language Concerns

BF_LC_1 There is little general information about blockchain for

certificates in Arabic. 5.11 1.628 .880

.935

BF_LC_2 There is little technical information about blockchain in

Arabic. 5.34 1.517 .880

The results of the CFA for the retained variables and items are presented in Table 52.

The model showed a good fit (CFI=0.966, NFI=0.950, TLI=0.954, RMSEA=0.082,

SRMR=0.077). All standardised regression coefficients were above 0.7 which represents

strong loadings. The convergent and discriminant validity of the data were estimated

based on the AVE scores. The AVE was estimated at 0.815, above the 0.5 threshold.

Therefore, convergent validity was established. Further, the AVE square root estimate of

0.903 was above the highest correlation of the variables (r = 0.445, lack of knowledge –

language concerns). Therefore, discriminant validity was also established.

Table 52: CFA for Barriers to Adoption Dimension and Items

Item Standardised Highest

Regression AVE √𝐀𝐕𝐄 Construct Model Fit

Weights Correlation

BF_LOK_1 .926

BF_LOK_2

BF_LOK_3

BF_LOK_4

BF_AWF_1

BF_AWF_2

BF_AWF_3

BF_AWF_4

BF_PAS_1

BF_PAS_2

.948

.935

.933

.734

.605

.944

.520

.854

.854 0.815 0.903 0.445

CFI=0.966

NFI=0.950

TLI=0.954

RMSEA=0.082

SRMR=0.077

BF_PAS_3 .971

BF_PAS_4 .944

BF_LC_1 .953

BF_LC_2 .923

7.3.6.6 Blockchain Adoption

Since the blockchain adoption intent dimension was represented by a single construct, a

reliability analysis only was performed on the data (Table 53). The data demonstrated

high levels of internal consistency for both individual items as measured by CITC and

the construct as measured by the Cronbach’s alpha. Therefore, all items were retained

for structural equation modelling analysis.

Table 53: Reliability Analysis for the Adoption Intent Dimension

Item Wording Mean St.D CITC Cronbach’s

alpha

Blockchain Adoption Intent

ADI_1 Our institution is considering blockchain adoption for

school certificates.

4.39 1.640 .690

.871 ADI_2 Our university is willing to adopt blockchain for school

certificates.

4.80 1.399 .837

ADI_3 Our institution will likely deploy blockchain for school

certificates in the nearby future.

5.03 1.419 .751

7.3.7 Multicollinearity Test

The final step in the preliminary analysis of the collected quantitative data was

multicollinearity analysis to identify potential issues with statistical significance of

independent constructs due to the high correlations among them. As recommended by

Hair et al. (2018) and Pallant (2010), tolerance levels ≥0.1 and VIF levels below 10

were taken as thresholds. Table 54 shows the results of a multicollinearity test

performed on the 16 refined independent variables explored in the study. With the

lowest tolerance factor of 0.314 and the highest VIF of 3.182 for the relative advantage

factor, all refined constructs demonstrated acceptable levels in terms of

multicollinearity.

Table 54: Multicollinearity Analysis of Independent Study Constructs

Tolerance VIF

Language Concerns 0.695 1.439

Lack of Knowledge 0.642 1.558

Peer Pressure 0.601 1.663

Complexity 0.579 1.726

Association with Finance 0.550 1.817

Observability 0.503 1.986

Government Support 0.459 2.180

Existing Regulations 0.457 2.190

Privacy and Security 0.456 2.193

Internal Resources 0.422 2.368

Trialability 0.413 2.422

Top Management Support 0.373 2.683

Education Quality 0.358 2.793

Compatibility 0.339 2.952

Employment Quality 0.335 2.982

Relative Advantage 0.314 3.182

7.4 Hypothesis Testing

The final model that included 5 dimensions represented by 16 independent factors and

the dependent variable adoption intent was tested for direct relationships. The observed

paths are displayed in Figure 26.

Figure 26: Path Analysis of the Study Model

*p<.05; **p<.01; ***p<.001; dashed line indicates no significant relationship

Variables:

RelAdv Relative Advantage

Compat Compatibility

Trial Trialability

Observ Observability

Compl Complexity

TopM Top Management Support

IntRes Internal Resources

ExReg External Regulations

GvtSup Government Support

PeerP Peer Pressure

EdImp Education Administration Improvements

EmImp Employment Improvements

LckKn Lack of Knowledge

AssFin Association with Finance Only

PrSec Privacy/Security Concerns

Lang Language Concerns

AdInt Intention to Adopt

e1 Error Factor

7.4.1 Technology Factors

The effect of the technology dimension variables on adoption is visualised in Figure 27:

Amos Results for Technology Factors. The model for technology dimension was

significant (R2 = .349, p < .001). Of the five factors in the technology dimension,

statistically significant relationships were observed for three (Table 55). The strongest

relationship was demonstrated by observability (β = .411, p < .001), followed by relative

advantage (β = .237, p = .043) and complexity (β = -.186, p = .027). Two factors did not

demonstrate statistically significant relationships: trialability (β = .226, p = .094) and

compatibility (β = -.009, p = .863).

Figure 27: Amos Results for Technology Factors

Table 55: Hypothesis Testing for the Technology Dimension Variables

Standardised

Relationship

Regression

S.E. t-value P

Hypothesis

Confirmed?

H1a: Relative Advantage -> Adoption Intent .237 .143 2.023 .043 Yes

H1b: Complexity -> Adoption Intent -.186 .128 -1.658 .027 Yes

H1c: Trialability -> Adoption Intent .226 .131 2.257 .094 No

H1d: Observability -> Adoption Intent .411 .069 5.616 <.001 Yes

H1e: Compatibility -> Adoption Intent -.009

R2 = .349, p < .001

.044 -0.172 .863 No

7.4.2 Organisational Factors

The results for the organisational factor dimension are displayed in Figure 28.

Figure 28: Amos Results for Organisational Factors

Both factors from the organisational factors dimension demonstrated positive,

statistically significant relationships to adoption intent: top management support (β =

0.374, p < .001) and institutional readiness (β = .411, p < .001). The results are

demonstrated in Table 56.

Table 56: Hypotheses Testing for the Organisation Dimension Variables

Standardised

Relationship S.E.

Regression

t-value P

Hypothesis

Confirmed?

H2a: Top Management Support -> Adoption

.374 .079

Intent

4.609 <.001 Yes

H2b: Institutional Readiness -> Adoption Intent .411 .079

R2 = .367, p < .001

4.964 <.001 Yes

7.4.3 Environmental Factors

The output from the Amos model for the environmental factors variables is presented in

Figure 29.

Figure 29: Amos Results for Environmental Factors

Two of the three external environment factors showed statistically significant

relationships to adoption intent: existing regulations (β = 0.312, p <.001) and

government support (β = 0.317, p<.001). However, the peer pressure factor was not

significantly related to adoption intent (β = .080, p = .231). The results are demonstrated

in Table 57.

Table 57: Hypotheses Testing for the External Factors Dimension Variables

Standardised

Relationship

Regression

S.E. t-value P

Hypothesis

Confirmed?

H3a: Existing Regulations -> Adoption Intent .312 .066 4.077 <.001 Yes

H3b: Government Support -> Adoption Intent .317 .099 3.586 <.001 Yes

H3c: Peer Pressure -> Adoption Intent .080

R2 = .367, p < .001

.070 1.198 .231 No

7.4.4 Quality Factors

Amos results for the quality factors are presented in Figure 30.

Figure 30: Amos Results for Quality Factors

Of the two quality factors, expected employment improvement demonstrated a positive,

statistically significant relationship to adoption intent (β=.602, p < .001). At the same

time, no significant relationship was observed for the education improvement variable

(β=.145, p = .091). The results are presented in Table 58.

Table 58: Hypotheses Testing for Quality Factors Dimension Variables

Standardised

Relationship S.E. t-value

Regression

Hypothesis

Confirmed?

H4a: Education Improvement -> Adoption Intent .145 .082 1.690 .091 No

H4b: Employment Improvement -> Adoption Intent .602 .102 6.288

R2 = .436, p < .001

<.001 Yes

7.4.5 Barriers to Adoption

Amos results for barriers to adoption are presented in Figure 31.

Figure 31: Amos Results for Barriers to Adoption

Of the barriers to adoption, three appeared to have a statistically significant influence on

adoption intent. As expected, all relationships were negative. Association with only

finance was the strongest factor (β = -0.494, p < .001), followed by privacy and security

concerns (β = -0.247, p < .001) and language concerns (β = -0.199, p = .002). The only

factor that did not show a statistically significant influence was lack of knowledge (β = -

0.062, p = .354). The results are presented in Table 59.

Table 59: Hypothesis Testing for Barriers to Adoption Dimension Variables

Relationship

Standardised

Regression S.E.

value

Hypothesis

Confirmed?

H5a: Lack of Knowledge -> Adoption Intent -.062 .058 -.928 .354 No

H5b: Association with Finance -> Adoption

Intent -.494 .071 6.785 <.001 Yes

H5c: Privacy and Security Concerns ->

Adoption Intent -.247 .048

<.001

3.822

Yes

H5d: Language Concerns -> Adoption Intent -.199 .055 3.036 .002 Yes

R2 = .230, p < .001

7.5 Chapter Summary

This chapter presented the results of the quantitative analysis of the collected survey

data. The total number of valid responses was deemed adequate for further analysis, and

the data displayed high levels of internal consistency and validity. Of the 16 hypotheses

formulated within the conceptual framework, 11 were confirmed by the results of the

SEM: 3 out of 5 for the technology dimension, 2 out of 2 for the organisational

dimension, 2 out of 3 for the environment dimension, 1 out of 2 for the quality

dimension, and 3 out of 4 for the barriers to adoption dimension. The next chapter offers

a comprehensive review and discussion of the study findings.

8 Discussion

8.1 Introduction

This chapter offers a comprehensive discussion of the study results. The results of the

qualitative and quantitative analyses are linked to the study goals and objectives: 1)

identifying positive and negative factors influencing blockchain adoption in Saudi HEIs;

2) developing and crystallising a framework of blockchain adoption; and 3) testing the

effect of factors/dimensions on blockchain adoption. Section 8.2 discusses the findings

with regard to the main research question posed in this study. Section 8.3 begins by

revisiting the research questions and main propositions to discuss their evolution as the

research progressed. This follows from the changes applied after the model presentation

and analysis in Chapter 6. Section 8.4 offers a comparative overview of the results in

Phase I and Phase II. Sections 8.5-8.9 discuss the findings in relation to the research

subquestions and hypotheses. The key relationships are identified and discussed in the

context of Saudi Arabia as well as the general propositions in the existing literature

related to blockchain adoption in education.

8.2 Main Research Question

The main research question in the study was:

How can we develop a holistic blockchain adoption model for Saudi HEIs?

The approach to developing a holistic actionable model successfully followed the design

science approach (Peffers, Tuunanen, Rothenberger, & Chatterjee, 2007). The model

was developed and evaluated in six steps:

1. Problem identification and solution value. A thorough literature review

enabled the existing knowledge gaps to be identified such as the absence of a

holistic framework for blockchain adoption designed specifically for Saudi

HEIs. Before this study’s inception, there were no empirically validated

blockchain adoption frameworks of such kind. Accordingly, the research

question was formulated and split into subquestions to deal with the complexity

more effectively.

2. Define the solution objectives Next, the objectives were formulated to provide

an actionable framework of blockchain adoption for Saudi HEI administrators.

Resources to meet the objectives were discussed: available theoretical and

empirical studies of adoption, industry experts’ opinions, and a large-scale

survey to confirm the role of specific model elements.

3. Design and development. The actual development of the model involved

several steps. First, the theoretical literature on blockchain adoption was

consulted to select the most appropriate theoretical foundations and determine

the main dimensions of the framework: technology, organisation, environment,

quality and barriers. Second, the selected theories were integrated to offer a

holistic adoption perspective. Third, the empirical literature was reviewed to

identify the factors within each dimension that showed a strong influence on

blockchain adoption in higher education.

4. Demonstration. The model was presented to a group of industry experts for

evaluation and analysis. In the course of the interviews, the experts defined the

most viable application of the proposed framework which is for

blockchainsupported school certificates. Accordingly, the initial model was

refined by integrating additional factors and removing some factors that the

experts considered less important in the context of Saudi HEIs.

5. Evaluation. The refined model was empirically tested and validated by

conducting a survey on a large group of higher education professionals in Saudi

Arabia. The evaluation confirmed the proposed model’s efficacy in explaining

the blockchain adoption process in Saudi HEIs.

6. Communication. The results of the model development and testing are

presented and thoroughly discussed in this thesis. A series of publications are

expected for further dissemination of the acquired knowledge. The process has

been already initiated by the submission of the literature review part of the thesis

to a peer reviewed journal and the qualitative analysis of the framework to a

conference.

8.3 Evolution of Research Subquestions and Objectives

As discussed in Chapter 2 of the current thesis, blockchain adoption in education is

lagging behind other industries such as finance, supply chain management and

healthcare. Nevertheless, the progress in research has been notable over the past few

years, and it allows strong parallels to be drawn to other industries in terms of adoption

factors.

Following this logic, the original framework developed for this study was based on the

technology adoption theories applied at the organisational level (Rogers E. , 2003;

Tornatzky & Fleischer, 1990) and available empirical literature covering the adoption of

blockchain in education (El Nokiti & Yusof, 2019; Fedorova & Skobleva, 2020;

Kosmarski, 2020; Ullah, Al-Rahmi, Alzahrani, Alfarraj, & Alblehai, 2020; Widjaja,

Cassandra, Widjaja, Prabowo, & Fernando, 2020). At the same time, recognising

contextual differences in the blockchain adoption process, the study refined the original

framework based on the interviews with the Saudi IT and education professionals.

Based on the interviews’ analyses, five dimensions comprising 16 factors in total were

retained as significant. The resulting framework, to the best knowledge of the author,

became the first of its kind to explain and predict blockchain adoption by Saudi HEIs.

The following research questions and hypotheses were formulated and investigated:

RQ1: Which technology factors influence blockchain adoption in Saudi HEIs?

H1a: blockchain’s relative advantage positively influences its adoption intent in Saudi

HEIs.

H1b: blockchain’s complexity negatively influences its adoption intent in Saudi HEIs.

H1c: blockchain’s trialability positively influences its adoption intent in Saudi HEIs.

H1d: blockchain’s observability positively influences its adoption intent in Saudi HEIs.

H1e: blockchain’s compatibility positively influences its adoption intent in Saudi

HEIs.

RQ2: Which organisational factors influence blockchain adoption in Saudi HEIs?

H2a: top management support positively influences blockchain adoption intent in

Saudi HEIs.

H2b: organisational readiness positively influences blockchain adoption intent in

Saudi HEIs.

RQ3: Which external factors influence blockchain adoption in Saudi HEIs?

H3a: existing regulations negatively influence blockchain adoption intent in Saudi

HEIs.

H3b: government support positively influences blockchain adoption intent in Saudi

HEIs.

H3c: peer pressure positively influences blockchain adoption intent in Saudi HEIs.

RQ4: Which quality factors influence blockchain adoption in Saudi HEIs?

H4a: Education improvement positively influences blockchain adoption intent in

Saudi HEIs.

H4b: Employment improvement positively influences blockchain adoption intent in

Saudi HEIs.

RQ5: What are the barriers to blockchain adoption in Saudi HEI?

H5a: lack of knowledge negatively influences blockchain adoption intent in Saudi

HEIs.

H5b: Association with finance negatively influences blockchain adoption intent in

Saudi HEIs.

H5c: Privacy and security concerns negatively influence blockchain adoption intent in

Saudi HEIs.

H5d: Language concerns negatively influence blockchain adoption intent in Saudi

HEIs.

Following is a discussion of the study results based on the research questions.

8.4 Comparison of the Qualitative and Quantitative Results

A model with a good predictive power should generally demonstrate similar to the

predicted results when tested on large populations. One of the main goals of this study

was to produce a working, reliable model of blockchain adoption in Saudi HEIs taking

into account the specifics of the context. Phase I of the research revised the initial model

and presented a finalised framework with 16 independent factors deemed significant for

blockchain adoption which were later empirically tested in Phase II on a large sample of

IT professionals and administrators in Saudi HEIs. Table 60 provides a comparison of

the predicted results developed during Phase I and the actual results obtained during

Phase II of the research. It can be seen that 11 out of 16 hypothesised relationships were

confirmed. On the one hand, this suggests that the developed model was relatively

robust in explaining blockchain adoption in Saudi HEIs. On the other hand, 5

invalidated predictions require further analysis. The next sections discuss the influence

of each factor on blockchain adoption individually and within their corresponding

dimensions.

Table 60: Comparison of Qualitative and Quantitative Results

8.5 The Role of Technology Factors

The role of technology factors in the blockchain adoption process is well recognised in

both the theoretical and empirical literature. Within the TOE framework, technology

innovation represents an array of novel technology solutions that could be beneficial for

a company (Lustenberger et al., 2021; Tornatzky et al., 1990). The DOI, in turn,

proposes that innovations contain five important characteristics that that make them

attractive for organisations: relative advantage, observability, complexity, trialability and

Research Questions and Hypotheses RQ1:

Which technology factors influence blockchain

adoption in Saudi HEIs?

Interviews* Survey Findings

H1a: Relative Advantage -> Adoption Intent Strong positive Supported Positive

H1b: Complexity -> Adoption Intent Moderate negative Supported Negative

H1c: Trialability -> Adoption Intent Moderate positive Not supported -

H1d: Observability -> Adoption Intent Moderate positive Supported Positive

H1e: Compatibility -> Adoption Intent RQ2:

Which organisational factors influence

blockchain adoption in Saudi HEIs?

Moderate positive Not supported -

H2a: Top Management Support -> Adoption

Intent Strong positive Supported Positive

H2b: Institutional Readiness -> Adoption Intent Strong positive Supported Positive

H2c: Organisational Size -> Adoption Intent

RQ3: What environmental factors influence

blockchain adoption in Saudi HEIs?

Path removed - -

H3a: Existing Regulations -> Adoption Intent Strong negative Supported Positive

H3b: Government Support -> Adoption Intent Moderate positive Supported Positive

H3c: Peer Pressure -> Adoption Intent RQ4:

Which quality factors influence blockchain

adoption in Saudi HEIs?

Moderate positive Not supported -

H4a: Education Improvement -> Adoption Intent Strong positive Not supported -

H4b: Employment Improvement -> Adoption

Intent Strong positive Supported Positive

Administration Improvement

RQ5: What are the barriers to blockchain

adoption in Saudi HEIs?

Path removed - -

H5a: Lack of Knowledge -> Adoption Intent Moderate negative Not supported -

H5b: Association with Finance -> Adoption

Intent Strong negative Supported Negative

H5c: Privacy and Security Concerns -> Adoption

Intent Moderate negative Supported Negative

H5d: Language Concerns -> Adoption Intent Moderate negative Supported Negative

Risk Avoidance -> Adoption Intent Path removed - -

Lack of Specialists -> Adoption Intent Path removed - -

Lack of Infrastructure -> Adoption Intent Path removed - -

Lack of General Readiness -> Adoption Intent Path removed - -

Lack of Visibility -> Adoption Intent Path removed - -

compatibility (Rogers, 2003). Recent reviews of blockchain adoption in various

industries have demonstrated that these factors are also important for the adoption of

blockchain at the organisational level (e.g., Clohessy & Acton, 2019; Clohessy et al.,

2020; Hartley et al., 2022). The results of the qualitative research conducted in the

context of the current study largely agreed with the general propositions about

technological factors reported in the literature. There was universal consensus among

the expert interviewees that all technological factors proposed by the DOI theory were

relevant for blockchain adoption in Saudi HEIs. The study, therefore, investigated the

influence of all five DOI factors.

8.5.1 The Influence of Relative Advantage

Within the DOI, relative advantage of a new technology is regarded as the strongest

predictor of its adoption (Lustenberger et al., 2021; Rogers, 2003). This was largely

confirmed by the results of the qualitative data analysis in this study. Relative advantage

was by far the most consistently and highly rated technology adoption factor by the

interviewees. This is logical because for organisations, the key consideration in adopting

a technology rests on whether it should replace the existing systems and whether it

would be a good decision from an investment perspective in the long run. In the case of

blockchain, however, the situation is more complex because it has been consistently

compared to a wide range of existing technologies, such as spreadsheets, ERP systems,

CRM, internal databases and security mechanisms (e.g., Hartley et al., 2022;

Kouhizadeh et al., 2021; van Hoek, 2019; Xu et al., 2021). This was also reflected in the

interviews, as the respondents spoke about blockchain’s advantage in relation to security

programs, trust-enabling technologies, administration programs, and digital education

applications. Therefore, blockchain was not juxtaposed against a single legacy

technology system it was expected to replace, but rather considered a groundbreaking

technology that can challenge a host of existing systems and applications. In fact, this is

confirmed by a wide range of blockchain existing and potential applications in

education as reported in the literature: from certificates to education platforms to

academic integrity and administration (Capece et al., 2020; Fedorova & Skobleva, 2020;

Jirgensons & Kapieniks, 2018; Kamisalic et al., 2020).

Despite the predicted strong influence on adoption in the literature and during the

interviews, the SEM analysis demonstrated only a moderate strength (β = .237, p

= .043). Therefore, while the hypothesis was confirmed, the influence of relative

advantage as a factor was not as strong as predicted. This result could be explained by

the specifics of blockchain applications in industry. The DOI proposes that the perceived

benefits of a technology innovation and its relative advantage depend strongly on the

innovation nature and the adopting population (Rogers, 2003). As such, different

industries will recognise blockchain’s relative advantage differently, based on their

specific needs and expectations (Carson et al., 2018). As noted earlier, blockchain

applications in education fall behind other industries such as finance, healthcare or

supply chain management to name a few (e.g., Alazab et al., 2021; Hasselgren et al.,

2020; Kulkarni & Patil, 2020). Whereas in those industries, there is a strong

understanding about the specific business processes that can benefit from blockchain

adoption, education institutions may still be in the process of discovering its advantages

in various application areas. Indeed, blockchain in education so far seems to be strongly

established only for diplomas with practical applications in other educational aspects

being in the development stages. As such, a weaker than expected impact of

blockchain’s relative advantage in the context of Saudi HEIs is explained.

8.5.2 The Influence of Compatibility

Compatibility is considered another positive factor in technology adoption within DOI

(Rogers, 2003). While compatibility is a technology feature, the theoretical treatment of

this aspect goes beyond connecting well to legacy systems. It is seen in a broader sense,

touching upon regulatory and technical demands as well as organisational goals and

objectives (Hartley et al., 2021; Lustenberger et al., 2021). It is not surprising then that

in this study, the respondents spoke of the importance of blockchain compatibility with

all aspects of HEI management: from services to administration. The survey results,

however, did not confirm the link between compatibility and blockchain adoption in

Saudi HEIs (β = -.009, p = .863), which requires further discussion.

It should be noted that some interviewees, in fact, did not consider compatibility as

influential in the blockchain adoption process. The key arguments seem to converge on

the idea that blockchain is inherently compatible with the existing systems used in Saudi

HEIs, and, as a result, seeking compatibility should not be a concern. This could be

explained by the well-developed and integrated cloud solutions in Saudi HEIs.

Blockchain researchers generally consider the cloud infrastructure as conducive to

blockchain adoption (e.g., Clohessy and Acton, 2019; Hartley et al., 2021; Orji et al.,

2020). Therefore, compatibility with the existing IT systems at the institutional level

may not be seen as a critical issue for Saudi HEIs. Further, existing research on

blockchain adoption suggests that for organisational users, a much larger concern rests

with nontechnological compatibility. Specifically, the issue of compatibility with the

existing privacy laws or data management standards typically overrides technology

compatibility concerns (Clohessy et al., 2020; Rauchs et al., 2019). Consequently, such

concerns were proven to be substantial barriers to blockchain adoption in Saudi HEIs as

discussed later on.

8.5.3 The Influence of Observability

The DOI proposes that the technologies whose positive effects can be observed more

easily will be adopted more readily (Rogers, 2003). This effect, however, could be less

pronounced with novel, less mature technologies like blockchain. This was emphasised

in a number of works on blockchain adoption and applications which noted the

technology’s relative infancy (Clohessy et al., 2020; Dobrovnik et al., 2018; Hartley et

al., 2021). Some authors also argued that the lack of observability leads to many

blockchain projects being shelved before they could show positive results (Rauchs et al.,

2019). Uncertainty regarding the observability effect on blockchain adoption was also

clearly traced during the interviews in this study. There were opinions both in support of

blockchain observability and against it which made it difficult to formulate the specific

degree or direction of that effect. It was assumed then that for the decision makers in

Saudi HEIs, the positive effects of blockchain should be tangible enough and related to

both internal operations and external effects to become influential.

Still, the results of the survey analysis showed that the effect of observability on

blockchain adoption intent was statistically significant and relatively strong (β = .411, p

< .001). Therefore, it was demonstrated that blockchain’s observability positively

influences the decision to adopt it. This result can be explained by the knowledge effect

of the research sample. As noted in Chapter 6, of the 10 interviewees, only 5 expressed a

deep understanding of the topic, whereas the other 5 acknowledged somewhat limited

knowledge. Consequently, there was no consistency among the interviewees regarding

blockchain adoption in education, including Saudi Arabia. Those with deeper

knowledge of the subject, in fact, spoke about the observability of the technology more

positively. In a similar manner, survey respondents who were more knowledgeable and

technically adept in blockchain technology could have recognised its potential for

education institutions better.

8.5.4 The Influence of Complexity

Technology complexity is considered a negative factor within DOI because the more

degree of effort that is required to understand and use technology, the harder it is to

adopt it (Rogers E. , 2003; Saberi, Kouhizadeh, & Sarkis, 2019). Over the course of the

interviews, it became apparent that complexity is, indeed, a negative feature of

blockchain which could slow its adoption in Saudi HEIs. The respondents

acknowledged such issues as a lack of blockchain understanding and technical expertise,

which would inevitably require additional training and/or hiring blockchain specialists

or consultants. This is largely in line with the existing literature that lists these issues

due to blockchain’s novelty (Clohessy & Acton, 2019; Falcone, Steelman, & Aloysius,

2021; Sternberg, Hofmann, & Roeck, 2021).

The results of the survey analysis confirmed the negative effect of complexity on

blockchain adoption intent in Saudi HEIs (β = -.186, p = .027), although the overall

negative effect was somewhat lower than expected. Perhaps, some explanation for this

can be attributed to good technology financing and well-developed IT departments in

Saudi HEIs. During the interviews, the respondents recognised the perceived complexity

of blockchain; however, they also argued that the stronger a HEI’s IT department is

believed to be, the lower the impact of complexity is expected to be. Indeed, existing

research demonstrated that strong IT departments and infrastructure help organisations

overcome blockchain complexity issues (Clohessy & Acton, 2019; Sternberg et al.,

2021). Therefore, the results of this study confirm the general propositions regarding

blockchain complexity but at the same time show how perceived complexity may be

countered.

8.5.5 The Influence of Trialability

In DOI, trialability is considered an important factor of adoption because it allows

organisations to test a technology at a low cost before deploying it (Rogers, 2003;

Rosenberg, 1982). In relation to blockchain, previous studies found that organisations

would postpone its implementation until the concept is well proven in practice for the

applications they sought (Lustenberger et al., 2021; Schmitt et al., 2019). Similar

arguments were provided by some interviewees who described standard procedures for

implementing new technologies in their HEIs with trials coming prior to

organisationwide deployment. However, the results of survey data analyses did not

confirm the relationship hypothesis (β = .226, p = .094). The effect of trialability,

therefore, was not present for blockchain adoption in Saudi HEIs, which requires further

discussion.

The lack of a trialability effect could be the result of routine innovation trials in Saudi

HEIs. As several interviewees acknowledged, their institutions periodically test

technology innovations which have the potential to benefit their organisations.

Blockchain then, would be considered another technology for such regular testing. In

fact, trialability is sometimes discarded from DOI-based adoption studies because trials

are usually imposed by organisations that consider potential technologies as beneficial

(Agi & Jha, 2022; Chong, Lin, Ooi, & Raman, 2009). In other words, whether an

innovation is easier or more difficult to test becomes irrelevant when an organisation is

keen on exploring it. This could be the case with blockchain in Saudi HEIs. The history

of cloud computing adoption also suggests that when a technology is seen as having

high degree of potential, Saudi HEIs do not consider trialability to be a serious factor.

8.6 The Role of Organisational Factors

The organisational dimension within the TOE framework incudes organisational

resources, structures and communication processes that influence decision making

regarding technology adoption (Baker, 2012; Tornatzky & Fleischer, 1990). The original

framework is rather flexible regarding what specific factors should be included in the

model. For example, a recent review of blockchain adoption studies by Clohessy et al.

(2020) identified 13 such factors. However, their strength of influence varied, and only

three were described as important: organisational readiness, top management support

and organisational size. These three factors were thoroughly investigated in this study in

the qualitative analysis stage which largely replicated the findings by Clohessy et al.

(2020), including a weaker effect of organisational size. The discussion of the findings

for each factor is provided below.

8.6.1 The Influence of Organisational Readiness

Organisational readiness is a comprehensive construct in TOE which covers a variety of

resources. Earlier research assessed organisational readiness from the perspective of

technological and financial resources (Iacovou, Benbasat, & Dexter, 1995). Recently,

organisational readiness, especially when it comes to blockchain, was expanded

substantially (e.g., Clohessy & Acton, 2019; Clohessy et al., 2020). The analysis of

qualitative data in this research largely confirmed the current literature on blockchain

adoption by considering organisational readiness as a combination of technological,

human and financial resources. Indeed, organisations that lack specialists, the

appropriate level of technology development and technology investments are often

considered incapable of adopting blockchain successfully (Post et al., 2018; Rauchs et

al., 2019).

The results of the survey analysis further supported the claim that organisational

readiness is a strong predictor of blockchain adoption (β = .411, p < .001). Therefore,

the results of this research aligned with the results of similar studies on blockchain

conducted recently. On a side note, some interviewees even dismissed the lack of one

aspect of organisational readiness by arguing that their institutions would be ready to

acquire the lacking resources if necessary. This may offer additional research avenues in

the future.

8.6.2 The Influence of Top Management Support

Top management support, along with organisational readiness, has been empirically

confirmed as a key factor in the adoption of innovation in general and blockchain

specifically (Clohessy, et al., 2020; Dong, et al., 2009; Duan, et al., 2020; Wong, et al.,

2020). The results of this study confirmed these findings. In the course of qualitative

data analysis, top management support was seen not only as an influential organisational

factor but also as “essential,” “necessary,” “significant” and “most important.” Further,

this factor was seen as the most important among the organisational dimension variables

in influencing blockchain adoption by Saudi HEIs.

In general, top management support in the adoption studies is considered from the

perspective of overcoming adoption barriers, creating a technology vision, and enabling

sufficient resource allocation for new innovations (Lustenberger et al., 2021). However,

it appears that in the case of Saudi HEIs, top managers are primarily considered the key

decision makers in adoption decisions per se. Due to stronger hierarchical structures in

Saudi organisations, top management support plays a stronger role. This, however, was

not confirmed by the survey results, since the effect of the top management support

factor was similar to that of organisational readiness (β = 0.374, p < .001). It is

reasonable to assume then that top management support and organisational readiness are

equally important when it comes to blockchain adoption in Saudi HEIs.

8.6.3 The Influence of Organisational Size

Organisational size is the often-considered construct in the organisational dimension of

the TOE framework. Recently, Clohessy et al. (2020) identified it as the third most

researched organisational variable in blockchain adoption research. However, the results

of the empirical research so far have been less conclusive than for organisational

readiness and top management support in the context of blockchain adoption. Some

studies found that larger organisations were better suited to blockchain adoption due to

resource availability, others found that smaller companies were more agile and elastic in

relation to embracing new innovations (Mendling et al., 2018; Tapscott & Tapscott,

2016; Wang et al., 2016). This uncertainty was traceable during the interviews in this

study as the respondents were split regarding which HEI size would be better suited for

blockchain adoption.

The decision to not consider organisational size as an organisational factor was also

made due to seeming agreement among the respondents that size was not a deterministic

term but rather a convenience measure of organisational resourcefulness. This follows

the logic of innovation adoption studies criticising organisational size as being less

specific and, therefore, meaningful than more concrete measures of resources required

for successful adoption (Baker, 2012). Resourcefulness, on the other hand, is reflected in

the organisational readiness variables which considered three major types of resources

for blockchain innovation. For these reasons, the decision to not consider the

organisation size – adoption intent path is justified. Still, the effect of organisational size

was measured as one of the control variables, and the relationship was not confirmed.

HEI size then was not proven as a significant factor in the blockchain adoption process

by Saudi HEIs.

8.7 The Role of Environmental Factors

The environmental context in the TOE framework includes factors beyond the firm’s

direct control that can influence innovation adoption (Baker, 2012; Tornatzky &

Fleischer, 1990). Market characteristics, industry characteristics, existing legal

frameworks and other factors have been considered for blockchain adoption (Clohessy

& Acton, 2019; Lustenberger et al., 2021; Zheng et al., 2018). However, environmental

factors are often regarded as contingent upon context and, therefore, specific to different

industries and countries. After the completion of the qualitative data analysis, this study

considered three environmental factors: existing regulations, government support and

peer pressure. The effect of existing regulations and government support on blockchain

adoption by Saudi HEIs was confirmed while the effect of peer pressure was not. The

results for each factor are discussed below.

8.7.1 The Influence of Existing Regulations

Government regulations represent an important area of influence on emerging

technologies, especially if such technologies are groundbreaking, redefining entire

industries (Piscini, Cotteleer, & Holdowsky, 2018). For organisations that consider

adopting such technologies, there is a large degree of uncertainty with regard to how

their use would be legally defined and regulated in the future. With regard to

blockchain, the current regulations do not always readily recognise how to govern it

from a legal perspective. Salmon and Myers (2019), for example, pointed out that legal

authorities would need to create better frameworks to regulate blockchain effectively.

The same concerns were also expressed by the interviewees in this study who

recognised that the existing regulations may not be able to duly cover various aspects of

blockchain adoption and applications in Saudi HEIs.

The significant influence of the existing regulations on blockchain adoption was

confirmed by the results of the survey data (β = 0.312, p <.001). Surprisingly, however,

the direction of the relationship was possible, thereby suggesting that the existing

regulations in fact supported blockchain adoption intent in Saudi HEIs. Several

explanations for this contradictory result are possible. First, when speaking of

regulations, the negative effect of existing regulations on blockchain adoption is often

attributed to uncertainty (e.g., Farooque et al., 2020; Lustenberger et al., 2021; Salmon

& Myers, 2019). Perhaps, when it comes to Saudi Arabia, the existing legislation is seen

as sufficiently mature, leaving less ambiguity about blockchain. Such optimism may

also be connected to a vibrant and conducive legal framework created for cloud

technologies in the country. Further, the respondents may have attributed such optimism

to the Vision 2030 policies which are very lenient for technology innovations of various

kinds. It would be, therefore, useful to examine the specific aspects of existing

regulations in Saudi Arabia and identify which ones provide such confidence for

potential organisational adopters of blockchain.

8.7.2 The Influence of Government Support

Governments are stakeholders in technology innovation processes, along with rival

firms, organisational customers and partners. The TOE recognises the role of

government as either positive or negative depending on whether certain technologies are

seen as prospective or threatening (Baker, 2012). With regards to blockchain,

government support is usually recognised as a positive driver of adoption because of the

technology’s newness (Chen et al., 2018; Farooque et al., 2020; Tapscott & Tapscott,

2016). The results of the qualitative analysis confirmed these literature findings. Further,

it was noted that such support increased during the COVID-19 pandemic due to an

increased focus on online technologies in education. This is also in line with the existing

literature which reported accelerated blockchain adoption in various sectors around the

globe (Shah, Shah, Tanwar, & Kumar, 2021; Yang, Zhang, & Shi, 2021). The results of

the survey analysis further confirmed the positive role of government support in

blockchain adoption by Saudi HEIs (β = 0.317, p<.001). Therefore, the study results

agreed with the existing theoretical and empirical postulates regarding the positive role

of government support on blockchain adoption by organisations.

8.7.3 The Influence of Peer Pressure

Competitors, along with the government, are traditionally regarded as the most

influential external stakeholders in the technology adoption literature (Iacovou,

Benbasat, & Dexter, 1995; Penttinen & Tuunainen, 2009). The extent of peer (also

known as mimetic) pressure regarding technology adoption arises from the perceived

success of rival organisations linked to new technology use. Accordingly, as the DOI

theory proposes, the innovators exert competitive pressures on other companies in their

industry to promote further technology diffusion (Rogers, 2003). Yet, the effect of peer

pressure on blockchain technology adoption is not clearly established empirically

(Iansiti & Lakhani, 2017; Kouhizadeh et al., 2021; Pilkington, 2016). The analysis of

qualitative data in this study also provided mixed conclusions about the effect of peer

pressure on blockchain adoption by Saudi HEIs. Whereas the supporters of the effect, in

line with the theory, pointed towards competitiveness blockchain provides, the

opponents believed that technology development in HEIs goes in accordance with

internal technology development plans. In the end, the survey results did not

demonstrate the significant effect of peer pressure (β = .080, p = .231).

The absence of a relationship between peer pressure and blockchain adoption In this

study can be explained in several ways. First, it is emphasised in both the theoretical and

empirical literature that peer pressure works best with technologies that are already

established (Baker, 2012; Hartley et al., 2021). Since blockchain remains in the early

application lifecycle, and not many HEIs in Saudi Arabia are using it, there is little

mimetic pressure on other HEIs to adopt it. Granted, in some cases, even technologies in

the early stages are pushed to adoption through partnerships or collaborative networks

that promote specific technology use (Lustenberger et al., 2021). However, this is not

the case with Saudi HEIs either because blockchain is still not widespread, and Saudi

HEIs can exercise independent policies in relation to technology choices. Finally, peer

pressure can be seen as only one of the factors in a large composition of market

dynamics (Clohessy et al., 2020). These dynamics may, in fact, challenge the existing

technology status quo and pressure organisations to adopt blockchain. However, this

could be a composition effect rather than a single factor effect of peer pressure.

Therefore, at least at this point in time, peer pressure alone is not sufficient to

substantially influence blockchain adoption among Saudi HEIs.

8.8 The Role of Education Quality Factors

While blockchain is a relatively new technology in the context of the education sector,

its potential contribution is rather high. Generally, it is recognised that innovative

technologies can improve education quality outcomes, especially in developing

countries (Duan et al. 2017, Xu et al. 2017, Farah et al. 2018, Williams, 2019). Existing

use cases identify at least a dozen working applications and many more potential

applications after integrating blockchain design platforms with the existing systems. As

determined in Chapter 2 of this thesis, enhancing education quality could occur in three

ways: 1) through an increased operational efficiency and, therefore, reduced cost of

education; 2) through new approaches to course delivery and enhanced learning

environments; and 3) through easily verifiable lifelong education credentials (AlHarthy

et al., 2019; Alammary et al., 2019; Bhaskar et al., 2020; Kosmarski, 2020).

Accordingly, this thesis identified education quality improvements in terms of

improving employment prospects for graduates, improving the quality of education

services and improving the quality of school administration.

8.8.1 The Influence of Intent to Reduce Unemployment

Reducing graduate unemployment is a context-specific issue for Saudi Arabia’s

education as the rate continues to hover around 28%, with over half of the unemployed

holding at least a bachelor’s degree (O'Neill, 2022). Improving access to the labour

market and empowering students are considered key indicators of education quality (Al-

Ramahi & Odeh, 2020; Harvey, 2006). Blockchain is considered to be a technology that

achieves this in two ways: 1) by offering an accelerated, easily verifiable system of

learning credentials and certificates; and 2) by creating job opportunities across the

blockchain system itself, which is related to engineering, software development,

education, administration and other related fields (Bucea-Manea-Tonis et al., 2021;

Guustaf et al., 2021; Hameed et al., 2019).

The results of this study confirmed the aforementioned assertions through both

qualitative and quantitative analyses. The majority of the interviewees agreed on the

strong effect of blockchain’s potential to reduce unemployment which, in turn, may

prompt its faster adoption. The inherent features of blockchain, such as immutability

and transparency, were commonly discussed as important factors in this regard.

Accordingly, the interviewees pointed to the ease and speed of the job application

process and the verification of student credentials using blockchain certificates. The

results of the survey data analysis further confirmed the link between unemployment

reduction and blockchain adoption intent in Saudi HEIs (β=.602, p < .001). Therefore,

the study results confirmed the existing propositions in the literature. On the other hand,

there was some degree of concern among the respondents about the ability of

blockchain to provide employment benefits to all students and not to early adopters

only. This kind of investigation was beyond the scope of this thesis but could,

nevertheless, offer an interesting avenue for further research.

8.8.2 The Influence of Perceived Quality of Service Improvements

Service quality in general refers to how well an organisation manages to deliver its

products and/or services to the customers (Prakash, 2019). Since HEIs are considered a

part of the service-providing industry, the issue of service quality is especially important

to them (Galeeva, 2016; Latif, Latif, Sahibzada, & Ullah, 2019). Technology is often

seen as a way to improve service quality in HEIs (Sugandi & Kurniawan, 2020). As

discussed in Chapter 2 of this thesis, blockchain improvements to higher education

service quality are expected to arise from organisation-centred improvements such as

efficiencies, administration and provision of courses and valuations. It is logical then

that the interview respondents saw improvements to quality of services arising from

blockchain as a combined effect on administrative and course delivery services. As

such, the qualitative analysis confirmed the proposition of service quality improvements

with blockchain but through a broader lens which included administration and learning

service provisions.

However, the results of the survey analysis did not support the link between

improvements to education service quality and blockchain adoption intent (β=.145, p

= .091). This may be explained by the fact that most of the advantages that blockchain is

expected to bring to education service quality remain in the conceptual stage at the

moment. Indeed, as the research shows, process automation, operational efficiencies,

data security and course delivery have been proposed in the form of models, but their

realisation in practice is still lacking (Alam & Benaida, 2020; Awaji et al., 2020;

Bhaskar et al., 2020; Mikroyannidis et al., 2018). Therefore, it may still be unclear what

particular aspects of service quality blockchain brings to Saudi HEIs. This could also be

indirectly assumed from the interviews, as the majority of the respondents did not offer

specific examples of service quality improvements that blockchain delivers or might

deliver to their institutions.

8.9 The Role of Barriers to Adoption

The presence of factors that obstruct or slow the adoption of innovative technologies is

well recognised in both DOI and TOE theories (Baker, 2012; Rogers, 2003; Tornatzky et

al., 1990). Being a relatively recent technology, blockchain is at the stage of

implementation where many barriers are likely to exist at the organisational level. The

empirical literature on blockchain adoption in education identified over a dozen barriers

attributed to the key dimensions of the TOE framework: technological, organisational

and environmental (e.g., Bhaskar et al., 2020; Ma and Fang, 2020; Yue et al., 2020).

However, a closer analysis of the barriers has led to an important observation that many

of them simply reflected the lack of enabling factors of adoption (for example, a lack of

resources and infrastructure) or reiterated the negative dimensional factors (for example,

technology complexity and regulations). Such factors were excluded from the analysis

of barriers to blockchain adoption by Saudi HEIs. Also, in the course of the interviews,

certain barriers were either mentioned rarely or not considered worthy of investigation.

In the end, four barrier factors were retained for the survey: the lack of knowledge about

blockchain, association with finance only, concerns regarding privacy and security and

language concerns. The effect of each of these factors on blockchain adoption is

discussed next.

8.9.1 Lack of Knowledge

The DOI posits that innovations must be understood before being adopted (Rogers,

2003). Innovations that are new and perceived complex will take more time to gain

followers. Existing research indicates that there is a general deficiency in blockchain

knowledge among the organisational decision makers, which makes it more difficult to

accept and adopt (Falcone et al., 2021; Post et al., 2018; Rauchs et al., 2019).

Consequently, a general lack of blockchain awareness in the educational sector was

reported by Fedorova and Skobleva (2020). The lack of knowledge about blockchain

also led to the perceptions of it being too complex (Castro and Au-Young-Oliveira,

2021; Liu & Zhu, 2021). In the course of the interviews in this study, similar issues were

identified. The respondents claimed that little was known about blockchain and its

applications to confidently start adopting it in Saudi HEIs.

However, the survey data analysis did not find a statistically significant relationship

between the lack of knowledge about blockchain and its adoption in Saudi HEIs (β =

0.062, p = .354). Therefore, the results of the study contradicted the generally

established negative link in the literature between these two factors (e.g., Falcone et al.,

2021; Post et al., 2018; Wang et al., 2019). One explanation for this could be that while

a lack of knowledge about blockchain is generally recognised in the Saudi educational

sector, it does not prevent universities from experimenting with it. Indeed, during the

interviews, none of the respondents, even those who claimed a drastic lack of

blockchain knowledge in Saudi HEIs, actually said that this would prevent their

institutions from adopting it. In contrast, the existing empirical studies view the lack of

knowledge as a strong preventive factor to adoption. For example, it was claimed that

the lack of knowledge causes organisations to postpone its adoption over concerns

regarding potential costs and the need for training (Clohessy & Acton, 2019; Sternberg

et al., 2021). However, as previously discussed, the majority of Saudi HEIs tend to

experiment with new technologies even if they lack substantial expertise in these

technologies. With this, they follow well-established technology strategies of their own.

Moreover, as reported by the interviewees, Saudi HEIs see no problem in acquiring the

necessary resources and specialists for the technologies that are seen to be valuable and

beneficial. For these reasons, the absence of the negative link between the lack of

knowledge about blockchain and its adoption could be justified.

8.9.2 Privacy and Security Concerns

Concerns regarding privacy and security with novel technologies are nothing new.

Because the security of data is of the outmost importance for organisations, the decision

makers prefer to rely on the established security mechanisms to ensure high levels of

user data safety. Due to its decentralised nature, blockchain is often referred to as a trust-

less mechanism (Pandey & Litoriya, 2021). Because it relies on a distributed network to

conduct and confirm transactions, no third parties are involved, and no single

centralized structure is necessary to its use. All transactions can be conducted from

different systems and devices connected to a network (Chen et al., 2018).

And yet, privacy and security concerns are common with blockchain applications. These

may be related to the absence of commonly established algorithms and related design

components for data safety across multiple blockchains (Holotiuk, Pisani, & Moormann,

2018; Spychiger, Tasca, & Tessone, 2021). In the context of higher education,

researchers found that the transparent nature of blockchain may be a perceived threat to

user privacy (e.g., Awaji et al., 2020; Ma & Fang, 2020; Pfeiffer et al., 2020; Raimundo

& Rosario, 2021). However, in the course of the interviews, it was revealed that the

primary reason for privacy and security concerns among the decision makers in the

context of Saudi HEIs was the lack of a complete understanding of how blockchain

technology works. The results of the survey analysis also confirmed the negative effect

of privacy and security concerns on blockchain adoption (β = -0.247, p < .001).

Therefore, while supporting the main proposition in the literature that privacy and

security concerns may slow blockchain adoption, this study identified a different

mechanism for this which is the lack of due understanding of technology. This suggests

that even though the lack of knowledge was not established as a direct barrier to

blockchain adoption in Saudi HEIs, it may have an indirect influence by enhancing

privacy and security concerns about blockchain.

8.9.3 Association with Finance Only

The association of blockchain with finance only is rarely if ever considered as a barrier

to its adoption in other industries. However, this factor emerged strongly in the course of

this study’s interviews. Specifically, the respondents talked about blockchain being

associated with cryptocurrencies and its perceived absence of useful applications for

education. This may not be a surprising result, however. Originally, the working

blockchain technology was implemented as a part of the bitcoin cryptocurrency

framework (Nakamoto, 2008). An incredible rally of cryptocurrency values, the creation

of cryptocurrency exchanges, and the legalisation of cryptocurrencies as viable payment

tools have unquestionably contributed to blockchain interest from a financial

perspective. However, blockchain applications outside of cryptocurrency, even

successful ones, remain much less exposed to the general public. This has been

recognised in most of the literature surveys on blockchain applications in education

(Bhaskar et al., 2020; Kamisalic et al., 2020; Ma & Fang, 2020; Raimundo & Rosario,

2021).

The results of the survey analysis confirmed the negative link between blockchain’s

association with finance only and its adoption in Saudi HEIs (β = -0.494, p < .001). In

fact, this was also the strongest barrier to adoption among those investigated. As such,

this study demonstrated that a unique factor like association with limited applications

could be detrimental to the blockchain adoption process. On a further note, it can be said

that such associations also arise from an insufficient understanding of the technology

and its principles of work. That said, the lack of knowledge about blockchain once again

could play an indirect role in impeding blockchain adoption by establishing

preconceived assertions of limited applications among the potential users.

8.9.4 Language Concerns

Language concerns emerged in this study as yet another unique barrier in the blockchain

adoption process. Unlike decision makers in English-speaking countries or where

English is commonly learned and used, the dominant language in Saudi Arabia is

Arabic. Accordingly, decisions regarding innovations are often supported by information

available in the native language. The majority of interviewees brought up their concerns

about the lack of information about blockchain in Arabic, which would in turn leave it

out of the usual technology considerations by the decision makers in Saudi HEIs.

Consequently, the main concern by the interviewees was that administrators would feel

uncomfortable implementing technologies without thorough descriptions available in

their native language. Further, it was argued that the language barrier in blockchain may

be a reason for the shortage of good blockchain specialists and, as a result, prevent its

timely adoption.

The results of the survey analysis confirmed the negative link between language

concerns and blockchain adoption in Saudi HEIs (β = -0.199, p = .002). It appears that a

language barrier can be a detrimental factor for blockchain adoption in Saudi HEIs.

While such a link appears somewhat unique in relation to blockchain adoption

specifically, language as a barrier to technology adoption has been explored in relation

to other innovations. Some examples in this regard include mobile payment services,

cloud technologies, online education platforms and telemedicine among others (Hiran,

2021; Leng, Gu, & Dalte, 2015; Otieno, Liyala, Odongo, & Abeka, 2016; Rahman &

Hoque, 2018). Further, similar observations were reported in the context of Saudi Arabia

for such technologies as online healthcare and learning management systems (Alenezi,

2021; Alodhayani, Hassounah, & Qadri, 2021). Therefore, the language barrier for

blockchain adoption confirmed in this study supports and further expands the existing

literature reports of its negative effect on technology adoption.

8.10 Chapter Summary

This chapter critically discussed the results of the study in the context of the available

literature on blockchain adoption both in the education sector and beyond. With 11 out

of 16 predicted relationships confirmed by the study outcomes, it can be asserted that

the proposed conceptual model had a good predictive power. In most cases, where the

relationships were confirmed, the study propositions coincided with the general

literature on the topic of blockchain adoption in the context of higher education. The

relationships that were not confirmed can be explained by a deeper analysis of the study

context. In addition, two specific relationships, namely the language barrier and

blockchain being associated with finance only are promising avenues for further

research. The next chapter provides the main conclusions following the study findings

and proposes theoretical and practical implications arising from the study results.

9 Conclusions and Recommendations

9.1 Introduction

This chapter summarises the key research findings and uses them to draw the main

conclusions and recommendations arising from this work. Section 9.2 draws the key

conclusions based on the study results. The conclusions are organised to demonstrate the

role and effect of each of the five dimensions of the proposed adoption model. Section

9.3 outlines the key theoretical implications of the study. Section 9.4 outlines the major

practical implications for HEI administrators and the higher education industry in Saudi

Arabia as a whole. Section 9.5 lists the major limitations of the study. Finally, Section

9.6 provides suggestions for future research.

9.2 Main Conclusions

To the best of the researcher’s knowledge, this study offered the first attempt to examine

blockchain adoption by Saudi HEIs using an integrated DOI-TOE framework. and

provided a comprehensive list of factors that could be influential in this process.

Importantly, a mixed methods research design applied in the study allowed the effect of

these factors to be measured and to also explore the underlying mechanisms of these

effects. In other words, the study answered the questions whether and why with regards

to the hypothesised relationships. This, in turn, offers good grounds for both theoretical

and empirical implications.

9.2.1 Technology Factors Relevant to Blockchain Adoption

This study explored the effect of five technology characteristics based on DOI (Rogers,

1995). From the results of the study, it is concluded that the three factors relevant for

blockchain adoption in Saudi HEIs are relative advantage, complexity and observability

(Table 61).

The underlying mechanism supporting the effect of relative advantage is blockchain’s

groundbreaking nature which seemingly allows it to replace or supplement a number of

existing technologies for the HEIs’ benefit. On the other hand, there is recognition that

many of the potential advantages still have to be realised in practice in the higher

education sector. The underlying mechanism supporting the negative effect of perceived

complexity on blockchain adoption is novelty: given the lack of information and use

cases, decision makers still perceive blockchain as rather difficult to understand and

implement. Finally, the underlying mechanism for observability effect on blockchain

adoption is the clarification of value. For decision makers, a higher degree of blockchain

visibility means a stronger understanding of how their institutions could benefit from its

adoption.

Table 61: Technology Factors Influencing Blockchain Adoption and the Underlying Mechanisms

Technology Factor Effect on Adoption Underlying Mechanism

Relative Advantage Positive Groundbreaking nature: blockchain

enables many technologies used by HEIs

to be replaced or supplemented.

Complexity Negative Technology novelty: perceptions of it

being difficult to understand and

implement.

Observability Positive Clarification of value: more visibility

enables a better understanding of

blockchain benefits.

9.2.2 Organisational Factors Relevant to Blockchain Adoption

This study explored the effect of three organisational factors within TOE (Tornatzky &

Fleischer, 1990). Based on the study results, it is concluded that top management

support and organisational readiness are influential for blockchain adoption in the

context of Saudi HEIs (Table 62). Further, the effect of both factors was similar in

strength thereby suggesting the equivalent importance of both. The underlying

mechanisms for top management support are usually seen through overcoming the

adoption barriers, creating a technology vision, and enabling sufficient resource

allocation for new innovations (Lustenberger et al., 2021). In the case of Saudi HEIs,

additional supporting mechanisms are strong vertical hierarchy and the top-bottom

decision process. With regard to organisational readiness, the supporting mechanisms

for blockchain adoption are the presence of a combination of technological, human and

financial resources.

Table 62: Organisational Factors Influencing Blockchain Adoption and the Underlying

Mechanisms

Organisational Factor

Top Management

Support

Effect on Adoption Underlying Mechanism

Positive Overcoming adoption barriers and

enabling resource allocation via strong

hierarchy and top-bottom decision

making.

Organisational Readiness Positive Presence of sufficient technology, human

and financial resources to support

adoption.

9.2.3 Environmental Factors Relevant to Blockchain Adoption

This study explored the effect of three environmental factors within TOE (Tornatzky &

Fleischer, 1990). Based on the study results, it is concluded that existing regulations and

government support are influential for blockchain adoption in the context of Saudi HEIs

(Table 63). The primary mechanisms responsible for the positive relationship between

existing regulations and blockchain adoption are context specific: 1) there is less

uncertainty regarding blockchain regulation; 2) there is generally a rather vibrant and

conducive legal framework for emergent technologies in the country; and 3) Vision

2030 policies create very lenient framework for blockchain adoption. As for government

support, focus on innovative online technologies in education and support is seen as the

primary driver.

Table 63: Environmental Factors Influencing Blockchain Adoption and the Underlying

Mechanisms

Environment Factor Effect on Adoption Underlying Mechanism

Existing regulations Positive Low legal uncertainty and favourable

Vision 2030 policies.

Government support Positive Focus on development of innovative

technologies in online education.

9.2.4 Education Quality Factors Relevant to Blockchain Adoption

Two education quality factors were considered to influence blockchain adoption.

Reducing student unemployment was found to be the only relevant factor for Saudi

HEIs. Once again, this factor is context-specific issue for Saudi Arabia’s education as

the rate continues to hover around 28%, with over half of the unemployed holding at

least a bachelor’s degree (O'Neill, 2022). Accordingly, two mechanisms that link a

potential reduction in graduates’ unemployment to blockchain adoption in Saudi HEIs

are: 1) establishing accelerated, easily verifiable system of learning credentials and

certificates; and 2) by creating job opportunities across the blockchain system itself in

such fields as engineering, software development, education, administration and related

fields.

9.2.5 Main Barriers to Blockchain Adoption

Four potential barriers to blockchain adoption were identified in this study. Based on the

study results, three are relevant to blockchain adoption in Saudi HEIs: privacy and

security concerns, association of blockchain with finance only, and language concerns

(Table 64). The underlying mechanism for privacy and security concerns, which

negatively affects the adoption of blockchain in Saudi HEIs, is a lack of a thorough

understanding of how blockchain technology works. In the same manner, it can be

concluded that the lack of such a thorough understanding plays an important role in the

negative relationship between blockchain’s perceived association with finance only and

its adoption in Saudi HEIs. Finally, the negative link between language concerns and

blockchain adoption in Saudi HEIs is underscored by the perceived lack of complete

information about blockchain in Arabic. This is another unique contextual factor

identified in this study.

Table 64: Barriers to Blockchain Adoption in Saudi HEIs

Barrier Effect on Adoption Underlying Mechanism

Privacy and

security concerns

Negative Lack of thorough

understanding of blockchain.

Association with finance

only

Negative Lack of thorough

understanding of blockchain.

Language concerns Negative Absence of sufficient information about

blockchain in Arabic.

9.3 Theoretical Implications

This study applied an enhanced DOI-TOE framework to investigate blockchain adoption

in Saudi Arabia HEIs. Both DOI and TOE offer solid theoretical foundations and

frameworks to study innovations as they diffuse through societies and populations. The

current research adds to the existing literature testing TOE extended models and the

more specific TOE-DOI to study blockchain adoption (e.g., Barnes & Xiao, 2019;

Clohessy et al., 2020; Ullah et al., 2020). The following theoretical contributions are

provided by the study:

-Confirmation of the Science Design approach for blockchain adoption

modelling: the Science Design approach for modelling blockchain adoption

process in higher education was successfully confirmed in this study.

-Confirmation of enhanced TOE-DOI framework viability: the results of the study

demonstrated that an enhanced TOE-DOI framework is a viable tool to study

blockchain in Saudi HEIs with the results of most hypotheses tests aligning with

the existing literature findings. As such, the study confirmed the application of

these theories in a new context.

-Importance of the contextual factors confirmed: the study demonstrated that

TOEDOI can be extended to account for unique context characteristics. On the

one hand, this confirmed the framework’s flexible nature; on the other hand, it

supplemented the existing theoretical knowledge by testing the key propositions

of TOE and DOI in a new model. As such, the study once again demonstrated

that the factors missing or not accounted for in either framework should be

added to the analysis for a more comprehensive view on adoption.

-Validity of the proposed model confirmed: the majority of the relationships

within the proposed framework of blockchain adoption were confirmed.

Therefore, the framework can be considered a good theoretically supported

foundation for the future studies of blockchain adoption in the context of Saudi

HEIs. It filled the existing gap in knowledge in the absence of such a framework.

-Validity of the blockchain adoption instrument confirmed: the study offers a

validated instrument to study adoption either through a combination of the

considered factors or test the effect of each dimension more thoroughly. Overall,

the researcher finds DOI and TOE excellent theoretical foundations to study

blockchain adoption at an organisational level in Saudi organisations and

specifically to explain the phenomenon of blockchain adoption in Saudi HEIs.

9.4 Practical Implications

The existing literature suggests that blockchain adoption in higher education is lagging

behind other sectors such as, for example, supply chain management, healthcare or

finance. This was also confirmed in the course of the interviews with the higher

education and IT professionals. As such, studies such as this could help increase the

overall awareness of the potential of blockchain in higher education and how it can be

realised in practice to benefit both HEIs in Saudi Arabia and the education sector as a

whole. The major practical implications arising from this study are:

-Offering a good initial understanding of the blockchain adoption process in

Saudi HEIs: The intense interest in blockchain technology arguably brings as

much excitement as speculation about its uses and applications in higher

education. As such, a good understanding of the topic is necessary for a prudent

approach to blockchain adoption by institutional decision makers. The

framework of adoption presented and tested in this study can serve as a good

starting point in this process.

-Offering the first-of-its-kind practical model of blockchain adoption in Saudi

HEIs: to the best knowledge of the author, this was the first study of its kind to

empirically explore the combinatory effect of technological, organisational,

environmental and quality factors on and barriers to blockchain adoption in

Saudi HEIs. It can serve as a foundation for developing a tool to guide HEI

organisations and, perhaps, the industry as a whole on whether, when and how

the adoption process should proceed.

-Model flexibility and modularity: the proposed blockchain adoption model

conveniently organises the important factors that impact blockchain adoption in

Saudi HEIs into several dimensions. The decision makers can assess the

influence of each dimension and the variables it comprises before making a

decision regarding blockchain adoption and during the implementation process.

-Relative importance of factors and dimensions: not all of the predicted

relationships were confirmed by the study results. However, the absence of

statistically significant relationships may not necessarily indicate the absence of

any effect whatsoever. It could be argued that the factors that did not exhibit a

relationship to blockchain adoption were simply regarded as less important in

the early stages of the blockchain adoption process, since this is the current stage

in Saudi HEIs. This could change as blockchain becomes widely diffused in the

industry. In fact, DOI (Rogers, 2003) assumes that certain technology attributes

are more important than others at different adoption stages of innovation

diffusion. For example, relative advantage and complexity could be more

important at the early stages of innovation diffusion when the technology is new,

and the decision makers must see its benefit clearly. At the later stages, however,

compatibility could become more important as organisations will need to make

cost-benefit decisions regarding the innovation implementation. The same logic

applies to TOE (Tornatzky & Fleischer, 1990). For example, the influence of

environmental factors could increase or decrease if blockchain regulations

become stricter or more lenient.

-A validated assessment tool for blockchain adoption and the factors that

underlie the process: while created specifically for blockchain applications for

school certificates, the tool can be used to determine the external and internal

influences on the blockchain adoption process so that targeted decisions can be

made to act on them.

9.5 Research Limitations

This study has a number of limitations which should be mentioned and taken into

account when considering the study results. These limitations are based on the study

geography, context and the chosen methodology:

-The study was geographically limited to HEIs in Saudi Arabia. The findings of

the study, therefore, should be treated with caution in other national contexts.

Similarly, the presented conceptual framework may need to be adjusted by

researchers in other countries, especially with regard to the context factors

presented and explored. Arguably, both the findings and the framework might be

more applicable in countries with similar educational structures,

technologyrelated jurisdictions, language and national culture. Specifically, these

are the Arab countries of the Gulf such as Kuwait, UAE, Bahrain, Oman and

Qatar. Still, it is expected that at least some of the factors identified within DOI

and TOE should be applicable to explore blockchain adoption in other countries’

contexts since the effects of the variables confirmed in this study are very similar

to studies conducted in other countries.

-The study focused on higher education sector specifically. Industry-specific

findings of this study should be approached with caution when extrapolated to

other sectors. The specifics of technology adoption may vary in different

industries and therefore some factors may be less relevant while new factors may

emerge. For example, this study introduced an industry-specific education

quality dimension which may be less important for other industries and sectors.

Further, while the factors of adoption within both DOI and TOE are considered

universal for all forms of innovations, they have demonstrated different degrees

of strength for various innovations. A logical decision for researchers who decide

to take this framework as a basis for other industry research would be to

determine the stage of technology adoption in question. A better match would be

those contexts where blockchain is relatively new, as is the case with blockchain

in the higher education sector.

-The study explored a specific blockchain application: school certificates. The

reason for this was explained in Section 6.6: following the interviews, it was

determined that blockchain certificates are the only viable application of the

technology in Saudi HEIs. However, the literature mentions other working and

prototype applications for blockchain in education: student identity management

and solutions, cryptopayments, novel learning platforms, intellectual property

protection, administrative cost reduction, institution accreditation and academic

publishing platforms (Capece, Ghiron, & Pasquale, 2020; Fedorova & Skobleva,

2020; Haugsbakken & Langseth, 2019; Jirgensons & Kapenieks, 2018;

Kamisalic, Turkanovic, Mrdovic, & Hericko, 2020). These applications may be

driven by other types of factors or the role of the explored factors in this study

may differ for them.

-The study had a cross-sectional design. The research was performed on

crosssectional data and represented a specific point in time. While demonstrating

the current state of the nature of the blockchain adoption phenomenon in Saudi

HEIs, it does not offer information on how the phenomenon develops. Further,

according to Rogers (2003), the degree of significance of the factors may change

as the diffusion progresses. The same is true for the TOE factors (Tornatzky &

Fleischer, 1990). Therefore, as the diffusion of blockchain progresses in higher

education, the presented framework may require re-visiting and re-evaluation.

-The study participants might not have had the best knowledge of blockchain.

Given the early stage of blockchain diffusion in Saudi HEIs, it is unlikely that all

the decision makers had a thorough knowledge of the technology and its

potential benefits. The interviews, for example, demonstrated that only half of

the participants had a good degree of knowledge about blockchain and could

confidently speak of its existing applications and benefits. It is quite possible

then that relatively large number of the participants of the large-scale survey did

not possess high degree of blockchain knowledge and applications in higher

education. As such, the strength of some factors in the framework may not be

fully indicative of the real state of matters related to blockchain adoption in

Saudi HEIs.

It can be seen that some of the limitations presented above arose from the study’s focus

while others were addressed to the best ability of the researcher. Therefore, the

described limitations are unlikely to diminish the quality of the research. Rather, some

of them present new avenues for research which, the researcher hopes, will be addressed

in future studies.

9.6 Future Research

Due to the exploratory nature of this study, a number of interesting and important

avenues for future research arise based on the study findings and limitations:

-Further model clarification: the relationships explored in this study may require

further analysis and confirmation. This may be done by drawing from larger

samples, clarifying factors’ operationalisations and investigating the

relationships in more detail.

-Further clarification of factors: for example, the role of organisational readiness

as a composite factor stimulating blockchain adoption in Saudi HEIs can be

clarified. This study treated organisational readiness as a combination of

financial, human and technology resources sufficient for adoption. However, this

may not be a comprehensive list of factors. Others may include alignment with

organisational strategy or competitive model or network incentive structures,

among others (Clohessy et al., 2020).

-Clarification of unexpected results: for example, the existing regulations

surprisingly demonstrated a positive impact on adoption in this study. The

respondents may have attributed such optimism to the Vision 2030 policies that

are very lenient in relation to technology innovations of various kinds. It would

be, therefore, useful to examine the specific aspects of existing regulations in

Saudi Arabia and identify which ones provide such confidence for potential

organisational adopters of blockchain.

-Clarification of the role of knowledge: while knowledge was not established as a

direct barrier to blockchain adoption in Saudi HEIs, its indirect effects could be

plausible given that, for example, many respondents did not understand the

underlying mechanisms of blockchain and, therefore, considered it a threat to

privacy and security. Likewise, the lack of knowledge about blockchain could

play an indirect role in impeding blockchain adoption by establishing

preconceived assertions of limited applications among the potential users.

-Exploration of additional individual and factors: As Awa et al. (2017) argued,

even in organisational settings, the onus of the decision making remains on

individuals. Their decision regarding technology could be somehow influenced

by attitudes towards it, their perceptions of how it could fit with certain tasks

related to HEI administration and service provision and to what extent they

believe blockchain would be easy to implement and use. As discussed, such

factors are available in TRA, TPB, TAM and TTF theories. Granted, adding

individual- and task-related factors could increase model complexity. On the

other hand, this could also offer new insights about blockchain adoption in

higher education. At the same time, this study demonstrated the importance of

contextual factors in blockchain adoption.

-Exploration of additional contextual factors. Despite a comprehensive list of

factors explored in this study, additional contextual variables could have been

inadvertently omitted. For example, the extent to which blockchain adoption

benefits students in terms of job opportunities, whether early adopters or not, can

depend on various contextual factors, including the level of implementation,

support systems in place such as graduates with a blockchain qualification, and

the job market's readiness for blockchain-verified credentials. Acknowledging

these contextual nuances can lead to a more informed decision making for HEIs’

management. Therefore, exploring additional contextual factors at various levels

could also help enhance the adoption framework.

-Testing the model in new settings: As previously mentioned, the framework

could fit national settings which are similar to Saudi Arabia in terms of culture,

structure of educational sector, language and technology related jurisdictions. It

would be useful to test these assumptions empirically. It could also be useful to

conduct comparative studies to identify which factors are stronger in countries

with differences in culture and higher education sectors. Further, this study

considered higher education institutions only. The model could be tested in other

educational institutions, such as K-12. Finally, the study did not distinguish

between different types of blockchain. The three major types of blockchain are

permissionless, permissioned and hybrid (Tapscott & Tapscott, 2016). Due to the

differences between the three, their adoption process and factors may differ as

well. This represents another potential research direction. Finally, research could

explore whether the model holds for different applications of blockchain in

education and whether the role of the factors varies in strength across them.

-New approaches to evaluate the model: the cross-sectional design of this study

allows the factors influencing blockchain adoption to be evaluated at this point

of time, under the existing conditions. The role and strengths of the factors

within the framework could change as blockchain diffuses further among the

HEIs. For example, some interview participants in this study argued about the

ability of blockchain to provide employment benefits to all students and not to

early adopters only. Some also argued that peer pressure could become more

significant as blockchain further penetrates the higher education sector and its

benefits become more visible. As such, it would be useful to re-visit the

framework at later stages of the adoption process as described within the DOI

(Rogers, 2003). Even better evidence could be provided by employing a

longitudinal research design to study blockchain adoption in Saudi HEIs which

could help observe changes in the factors’ dynamics as the occur.

9.7 Chapter Summary

This chapter provided the major conclusions arising from the research, drew

implications from the findings and proposed directions for future research. The mixed

research method applied in the study allowed both the relationships between various

factors of blockchain adoption and its adoption in Saudi HEIs and the mechanisms

supporting those relationships to be uncovered. These were summarised in this chapter.

Further, the study demonstrated the applicability of the key theoretical frameworks, DOI

and TOE in combination to explore blockchain adoption in the new context of Saudi

HEIs. The developed framework that distinguished technology, organisational,

environment and quality factors as well as barriers could serve as a basis for future

research on blockchain adoption. Since the framework is both theoretically supported

and flexible, it can be modified and adjusted to meet the requirements of other contexts.

The results of the study also offer a good initial map for decision makers outlining the

important internal and external factors that matter in the blockchain adoption process.

The chapter outlined several important limitations that should be considered alongside

the research findings. These are based on the study geography, context and the chosen

methodology. Considering these limitations, several directions for future research were

identified: 1) applying the study framework in different national and sectoral contexts;

2) refining the framework by clarifying and adding/removing certain factors; 3)

clarifying the role of the existing factors; and 4) observing possible changes to the factor

roles at different stages of the diffusion process.

Students also viewed