An Examination of Small Businesses Propensity to Adopt Cloud-Computing

CSE 546 - Cloud Computing

Arizona State University

Introduction

Small business employment and production are essential for the economic health

of the United States (Small Business Administration [SBA], 2009). Information

technology (IT) innovation has been shown to foster sustainable business competitive

advantage and economic viability (McAfee & Brynjolfsson, 2008; Oh & Pinsonneault,

2007; Porter & Millar, 1985). Cloud computing, the elastic and ubiquitous provisioning

of IT services via the Internet, was envisioned as an emerging innovation enabling

enterprise strategies via a novel business-computing paradigm to achieve competitive

advantage (Cheng, 2010; Gens et al., 2010; Smith, 2009; Wyld, 2009).

In this doctoral study, small business leaders’ perceptions of cloud-computing

innovation and their propensity to adopt this emerging technology were examined.

Background information is presented as a foundation for the economic importance and

volatility of U.S. small businesses. IT innovation is discussed as a basis for contributing

to competitive advantage. Cloud computing was described as an innovative

businesscomputing paradigm delivering economic value via global IT scale and scope

(Mell & Grance, 2010; Truong, 2010). Based on the predictive ability derived from the

diffusion of innovations theory, specific research questions were addressed by testing

hypotheses that examined the relationship of perceived innovation attributes and small

businesses’ propensity to adopt emerging cloud-computing technology. A review of the

literature is presented to support the significance of this research, which was framed in

addressing the gap in understanding related to small businesses’ proclivity to use cloud-

computing innovation as a means of obtaining economic value.

Background of the Problem

The subject area of interest for this doctoral research project was the propensity of

small business leaders to adopt emerging cloud-computing innovation. The

conceptualization of cloud-computing innovation was inspired in the genesis of the

Internet (Buyya, Yeo, Venugopal, Broberg, & Brandic, 2009). Envisaged as utility-like,

cloud computing was featured as the ubiquitous, on-demand, self-service, elastically

scalable, pay-for-use provisioning of information technologies via the Internet (Cheng,

2010; Mell & Grance, 2010; Smith, 2009). Cloud-computing innovation was viewed as

an emerging economy-of-scale leveraged commoditization of IT services including

hardware, software, networking, or a combined platform of services together with its

undergirding human capital IT acumen.

The long-term health and vitality of small businesses have been essential for

economic prosperity and enjoyment of personal freedoms (Anderson, 2009; Michael &

Pearce, 2009; Shackelford, 2009; SBA, 2009). The volatility and high rate of

discontinuances of small businesses have become an economic threat for community and

government interests (Bates, 2005; SBA, 2009). The lack of sustainability in small

businesses threatens the viability of the United States’ free market economic system.

Established to preserve free competitive enterprise, the SBA (2009) has

recognized the critical importance of small businesses to U.S. economic strength as well

as their role in the global marketplace. For the purposes of this doctoral study, the

meaning of a small business was defined as a firm employing fewer than 500 employees

and generating annual revenues of $50 million or less, which was consistent with the

SBA’s definition. In the United States, small businesses have been reported as creating

the greater part of the gross domestic product while comprising nearly all employers and

employing the majority of the private sector workforce (Ibrahim, Angelidis, & Parsa,

2008; SBA, 2009).

An examination of the research revealed that innovation is essential for the

economic benefits of sustainable business vitality (Braganza, Awazu, & Desouza, 2009;

Elmore, 2007; Estrin, 2009; Mangelsdorf, 2009; Michael & Pearce, 2009). Innovation

has been shown to yield competitive advantage via unique business process capabilities

(Chalhoub, 2010; Hoerl & Gardner, 2010; Latzer, 2009), novel products and services for

existing markets (Dew, Sarasvathy, Read, & Wiltbank, 2008), and disruptive creations of

new markets (Carayannopoulos, 2009; Dewald, & Bowen, 2010; Hagel, Brown, &

Davison, 2008). Specifically, investments in IT innovation and IT dependent strategies

have been correlated with economic performance and competitive advantage (Latzer,

2009; McAfee & Brynjolfsson, 2008; Oh & Pinsonneault, 2007; Qureshil, Kamal, &

Wolcott, 2009).

Unprecedented acceleration in competition triggered by Internet globalization

amid the business volatility induced by recent economic turmoil has hastened the

mandate for IT agility and efficiency at lower costs (Liao, Welsch, & Moutray, 2008;

McAfee & Brynjolfsson, 2008; Renski, 2009). IT innovation has served as a catalyst for

improving business performance and economic value (Dibrell, Davis, & Craig, 2008;

Latzer, 2009; Porter & Millar, 1985). Anticipating delivery of economic value essential

for business, cloud computing has been portrayed as a novel business computing

paradigm extending IT flexibility and improving IT capabilities via ubiquitous

provisioning of IT services (Katzan, 2008; Sultan, 2010). The pay-for-use aspect of the

cloud-computing business paradigm was anticipated to mitigate capital-intensive IT costs,

thereby benefiting capital-constrained small businesses (Marston, Li,

Bandyopadhyay, Zhang, & Ghalsasi, 2010).

Diffusion of innovation (DOI) and technology acceptance model (TAM) theories

have been used in research to understand the diffusion and acceptance of innovation

among members of society (Davis, 1989; Rogers, 2003). Although the DOI research

tradition has its genesis in diffusion of agricultural innovation inquiry dating back to the

early 1960s (Rogers, 2003), there has been extensive DOI research related to diffusion of

IT and Internet innovations (Andrés, Cuberes, Diouf, & Serebrisky, 2010; Rogers, 2003;

Valier, McCarthy, & Aronson, 2008). Similarly, TAM has been used extensively as a

theoretical framework for studying the acceptance of IT technologies based on ease of

use and usefulness measures (Azadegan & Teich, 2010; Davis, 1989; Venkatesh, Morris,

Davis, & Davis, 2003). Rooted in DOI and TAM theory, Valier et al. (2008) developed a

predictive model for researching the adoption of IT innovation during its prediffusion

stage. Valier et al.’s predictive DOI model undergirds the theoretical framework applied

to this inquiry for examining the relationship between cloud-computing innovation and its

adoption by small businesses.

The ability for small businesses to finance innovation has been a serious

consideration in its use and economic viability (Robb & Robinson, 2009; Schwienbacher,

2007). Entrepreneurial financing often has been restricted to bootstrap financing

techniques and limited debt financing whereas only a small fraction of small businesses

were able to secure adequate equity financing (Barringer & Ireland, 2010; Paul, Whittam,

& Wyper, 2007). Access to financial capital has been a factor influencing small

businesses’ investment in innovation and their strategies for growth. Technology

innovations such as the Internet have mitigated capital barriers that previously restricted

small business global competition, thus reducing the socioeconomic gaps often attributed

to the diffusion of innovation and the spread of new ideas (Rogers, 2003).

Alleviating the capital investment requirements of large-scale IT capability, cloud-

computing innovation has operationalized these front-loaded costs via pay-for-use global

rapid provisioning of seemingly infinite on-demand IT services creating a novel business-

computing model (Katzan, 2008, 2010; Marston et al., 2010; Mell & Grance, 2010). The

emergence of disruptive business model innovations has contributed to the rapidly

changing business environments (Dewald & Bowen, 2010). The cloudcomputing

business paradigm was perceived as a radical disruptive innovation supplanting extant

business computing models’ facilitation of competitive advantage. Early adoption of

disruptive innovation often has garnered a leveraged marketplace opportunity to protect

or unseat extant competitive advantage (Dew et al., 2008).

Large corporations and public sector organizations have been targeted as early

adopters of emerging cloud-computing innovation purposing to yield greater competitive

advantage and economic value (Braude, 2008; Davis, 2009; Gens et al., 2010; Mell &

Grance, 2009; Smith, 2009; Wyld, 2009). The methods for past small business survival

have undergone a transformation as small businesses have capitalized on emerging

technology innovations (Carayannopoulos, 2009) to compete in an increasingly global

economy. In a new era of competitive survival, early adoption of cloud-computing

innovation was anticipated to provide the opportunity for small businesses to capitalize

on this novel business-computing paradigm and to thrive in the changing global and

virtual business environment. With the exception of limited utility functions such as

hosted e-mail services (Schadler, 2009), a review of the research literature has revealed

that small businesses have not been early adopters of emerging cloud-computing

innovation (Truong, 2010; Wyld, 2009).

Scarcity of information related to the penchant of small businesses to adopt

cloudcomputing innovations was considered problematic because it is the sort of insight

useful for establishing policies and other positioning influences to foster improved small

business sustainability and economic vitality. Acquiring a better understanding of small

business leaders’ propensity to adopt cloud-computing innovations could help service

providers improve positioning of this emerging technology for greater small business

benefit and economic value. This scant understanding was addressed in this study by

examining small business leaders’ perceived attributes of cloud-computing innovation

and their propensity to adopt emerging cloud-computing technology.

Problem Statement

Small businesses that do not embrace cloud computing forfeit the benefits

attributed to its novel business-computing paradigm resulting in a competitive

disadvantage that further threatens small businesses’ sustainability and economic

contribution (Armbrust et al., 2010, Marston et al., 2010). The general problem targeted

was small businesses’ slow embrace of technology innovations for greater economic and

societal contributions. The specific problem researched in this study was the limited

early adoption of cloud-computing innovation by small businesses for improved

economic value. Adapting a resource-based view of IT to cloud computing, Truong

(2010) reported that less than 2% of small businesses have employed a limited form of

newly emerging cloud computing and further recognized the need to understand small

businesses’ potential use of cloud-computing technology. Modeling cloud-computing

business macroeconomics, Etro (2009) forecasted improved small business economic

contribution by simulating approximately 430,000 firm creations during a 5-year period

using a rapid cloud computing adoption rate, which was more than five times the number

of firm creations using a slow adoption rate.

Purpose Statement

The purpose for this quantitative study was to examine the relationship between

small business leaders’ perception of cloud-computing attributes and their propensity to

adopt this emerging innovation. Applying predictive DOI theory, this objective was

accomplished by collecting information from Arizona small business IT decision-makers

through a previously tested online survey instrument that was designed to assess the

adoption of IT during the prediffusion stage of innovation. To ensure an adequate

response size, a convenience sample of Arizona small businesses was drawn from an

entire e-mail list maintained by the Arizona Commerce Authority (ACA).

Constructs measuring the perceived attributes of cloud computing, which included

the independent variables compatibility, complexity, observability, relative advantage,

results demonstrable, trialability, and voluntariness, as well as the dependent variable

intent to use cloud computing, were collected via the survey instrument. The data

collected were analyzed using SPSS statistics to assess the relationship between small

business leaders’ perception of cloud-computing attributes and their intent to use this

emerging technology. Based on a nonexperimental cross-sectional survey design, this

research was not intended as an investigation of causation.

A better understanding of small business leaders’ propensity to adopt cloud

computing was envisaged to guide positioning of this emerging technology to foster rapid

adoption translating into increased entrepreneurial startups and stimulating small business

growth. Transcending economic value, contributions to social change were projected by

potentially guiding policy practices for developing employee skills, initiating private and

public sector collaboration, and anticipating the environmental consequences of

computing.

Nature of the Study

A quantitative research method was elected for this doctoral research project.

Rooted in a postpositivist worldview of determinism (Stacey, 2007), this inquiry was

undertaken to understand the relationship between small businesses’ perception of

cloudcomputing innovation attributes and their intent to adopt this emerging IT

innovation. Beyond advancing prediffusion stage DOI research and providing a

steppingstone for continued research linking cloud-computing technology to small

business competitive advantage, the immediate anticipated benefit of this understanding

was the potential for guiding the establishment of policies and other positioning

influences related to cloudcomputing innovation for more rapid adoption by small

businesses. Targeting this outcome, study hypotheses were stipulated to address specific

research questions based on Valier et al.’s (2008) theoretical model related to the adoption

of IT during the prediffusion stage of the innovation process.

Quantitative analysis was performed using SPSS’s statistical functionality

consisting of bivariate linear regressions and multiple regression computations. The

statistical functions performed were based on one dependent variable, the small business

leaders’ intent to use, and seven independent variables representing the perceived

innovation attributes: relative advantage, compatibility, complexity, observability,

trialability, results demonstrable, and voluntariness. The quantitative research method

was well suited for deductively testing objective theories by empirically measuring and

statistically analyzing variables to understand their relationships (Campbell & Stanley,

1963; Creswell, 2009; Onwuegbuzie, Johnson, & Collins, 2009).

Conversely, qualitative research methods are generally exploratory in nature,

inductively seeking to construe the meaning about a phenomenon derived from

participants’ vantage (Creswell, 2009; Onwuegbuzie et al., 2009; Yin, 2009). In

particular, Rogers (2003) reported that qualitative case studies have been effective in

exploring postdictive DOI social system and change agent influences. Mixed methods

research have been employed in other studies to combine the interdisciplinary strengths

of deductive and inductive practices attributed to quantitative and qualitative

epistemologies, respectively (Cassell, Buehring, Symon, & Johnson, 2006).

The inductive, open-ended exploratory aspect of the qualitative research method

was not congruent with this doctoral study’s predictive DOI research purpose, which was

proposed to examine deductively the relationship between the prediffusion attributes of

cloud-computing innovation and the intent of small businesses to adopt cloud-computing

technology. Moreover, the added complexity of mixed methods research extended

beyond the scope of this doctoral study’s purpose. Neither qualitative nor mixed methods

research were as well suited for this research study’s bounded scope as a quantitative

approach.

This research project was fashioned from a traditional quantitative cross-sectional

survey design strategy that included four key elements: literature-based theoretical

hypotheses and study variables, population sampling criteria, a single point of data

acquisition via a survey instrument, and statistical data analysis and interpretation

(Creswell, 2009). Although the conceptualization of cloud computing has its genesis in

the framing of the Internet, cloud computing was viewed as a newly emerging IT arena

(Buyya et al., 2009; Katzan, 2010; Louridas, 2010; Truong, 2010). The rationale for

using a cross-sectional survey design approach was anchored in providing statistical

inferences from measurement data collected at a specific point versus using multiple

survey results collected over a period.

The quantitative survey design approach was chosen over a structured interview

strategy to measure more efficiently and cost-effectively the requisite sample size in a

timely manner (Bardhan, 2007; Creswell, 2009). Furthermore, this cross-sectional

research design consisted of a web-based, self-administered data acquisition process.

Web-based survey tools have been employed in numerous IT-related surveys, offering

benefits over traditional survey practices (Oke, Burke, & Myers, 2007, Maronick, 2009).

The use of a web-based survey tool was preferred as responses were received

more quickly than with traditional mail methods while mitigating data entry errors

encountered in recording hardcopy survey results. The use of an online survey tool was

calculated to be lower in cost than the traditional mail methods because printing, mailing,

and handling expenses are mitigated. Additionally, the use of a web-based survey tool

benefited the participant by providing participant anonymity while capturing necessary

informed consent, eliminating manual response remittance activities, and assimilating

with modern experiences in electronic communication methods.

Arizona small businesses numbered approximately 106,800 firms in 2008 (U.S.

Census Bureau, 2008). Based on the U.S. Census Bureau 19 industry-sector

categorization, the top four numbering sectors, including construction, professional

together with scientific and technical, health care and social assistance, and retail trade,

comprised slightly over half the total number of Arizona small business. The study

population of Arizona small businesses was derived from ACA’s estimated 4,500

crossindustry sector e-mail addresses. Anticipating a small sample response rate of 3%,

the number of survey participants was expected to include around 135 senior IT

deployment decision-makers from Arizona small businesses. A study sample size of at

least 117 participants was needed based on a priori goodness-of-fit G*Power calculation

(Faul, Erdfelder, Buchner, & Lang, 2009). Based on the sampling method, cross-

sectional survey, and other study elections, this research project was considered a

nonexperimental quantitative design and was not intended as an investigation of

causation. Additional information elaborating on the specific details related to this

study’s research method and design is presented in Section 2.

Research Questions

To address the problem statement and to achieve the objectives of this study’s

purpose, given small business volatility and economic importance, the overarching

research question guiding this study was, “What is the propensity of small businesses to

embrace the cloud computing novel business-computing paradigm?” The relationship

between each of the independent variables composing the perceived attributes of

cloudcomputing innovation and the dependent variable intent to use cloud-computing

technology were addressed by the following specific research questions (RQ).

RQ1: To what extent does relative advantage, a perceived attribute of innovation,

relate to small business leaders’ intent to use cloud-computing services?

RQ2: To what extent does compatibility, a perceived attribute of innovation, relate

to small business leaders’ intent to use cloud-computing services?

RQ3: To what extent does complexity, a perceived attribute of innovation, relate

to small business leaders’ intent to use cloud-computing services?

RQ4: To what extent does observability, a perceived attribute of innovation, relate

to small business leaders’ intent to use cloud-computing services?

RQ5: To what extent does trialability, a perceived attribute of innovation, relate to

small business leaders’ intent to use cloud-computing services?

RQ6: To what extent does results demonstrable, a perceived attribute of

innovation, relate to small business leaders’ intent to use cloud-computing services?

RQ7: To what extent does voluntariness, a perceived attribute of innovation, relate

to small business leaders’ intent to use cloud-computing services?

RQ8: Which has a stronger relationship to small business leaders’ intent to use

cloud-computing services the perceived attribute of innovation relative advantage or

voluntariness?

The study focus, which was delineated by the research questions, was honed

further via the hypotheses stipulated in the following section.

Hypotheses

Congruent with the postpositivist deterministic research paradigm, the hypotheses

were a priori assumptions that were tested using statistical procedures as described in

Section 2 to determine the existence, strength, and direction of the study’s variable

correlations (Creswell, 2009). An entire list of the survey questions is presented in

Appendix A. Additionally, each survey question’s contribution to the meaning of the

variable constructs assessed via the hypotheses is identified in the Operational

Definitions of Variables section. The following are the hypotheses related to the research

questions assessed by this doctoral study.

Hypothesis 1

The first hypothesis (H1) was stated theorizing that, during the prediffusion stage,

the higher the level small business leaders perceived the relative advantage of cloud

computing the greater their propensity to adopt the innovation.

H10: No correlation exists between relative advantage and intent to use cloud

computing.

H1a: Relative advantage is correlated with intent to use cloud computing.

Hypothesis 2

The second hypothesis (H2) was stated theorizing that, during the prediffusion

stage, the higher the level small business leaders perceived the compatibility of cloud

computing with extant factors the greater their propensity to adopt the innovation.

H20: No correlation exists between compatibility and intent to use cloud

computing.

H2a: Compatibility is correlated with intent to use cloud computing.

Hypothesis 3

The third hypothesis (H3) was stated theorizing that, during the prediffusion stage,

the lower the level small business leaders perceived the complexity of cloud computing

the greater their propensity to adopt the innovation.

H30: No correlation exists between complexity and intent to use cloud computing.

H3a: Complexity is negatively correlated with intent to use cloud computing.

Hypothesis 4

The fourth hypothesis (H4) was stated theorizing that, during the prediffusion

stage, the higher the level small business leaders perceived the observability of cloud

computing the greater their propensity to adopt the innovation.

H40: No correlation exists between observability and intent to use cloud

computing.

H4a: Observability is correlated with intent to use cloud computing.

Hypothesis 5

The fifth hypothesis (H5) was stated theorizing that, during the prediffusion stage,

the higher the level small business leaders perceived the trialability of cloud computing

the greater their propensity to adopt the innovation.

H50: No correlation exists between trialability and intent to use cloud computing.

H5a: Trialability is correlated with intent to use cloud computing.

Hypothesis 6

The sixth hypothesis (H6) was stated theorizing that, during the prediffusion

stage, the higher the level small business leader perceived the results demonstrable of

cloud computing the greater their propensity to adopt the innovation.

H60: No correlation exists between results demonstrable and intent to use cloud

computing.

H6a: Results demonstrable is correlated with intent to use cloud computing.

Hypothesis 7

The seventh hypothesis (H7) was stated theorizing that, during the prediffusion

stage, the more mandatory (i.e., less voluntary) small business leaders perceived the

voluntariness of cloud computing the greater their propensity to adopt the innovation.

H70: No correlation exists between voluntariness and intent to use cloud

computing.

H7a: Voluntariness is negatively correlated with intent to use cloud computing.

Hypothesis 8

The eighth hypothesis (H8) was stated theorizing that, during the prediffusion

stage, small business leaders’ perceived relative advantage of cloud computing was more

strongly related to their propensity to adopt the innovation than the strength of

relationship between their perceived voluntariness and propensity to adopt cloud

computing.

H80: Voluntariness has an equal or greater correlation with intent to use cloud

computing than does relative advantage.

H8a: Relative advantage has a higher correlation with intent to use cloud

computing than does voluntariness.

Theoretical Framework

The theoretical framework for this study was based predominately on the DOI

theory. Commencing with his doctoral dissertation presentation in 1957, Everett M.

Rogers has championed DOI research, culminating in his internationally acclaimed work

Diffusion of Innovation, which had its fifth edition publishing in 2003. Rogers’s (2003)

research findings and generalizations related to DOI have been used as the theoretical

framework for a multitude of studies examining the process of innovation communication

via channels to members of social systems over time. Although the genesis of DOI grew

out of various research traditions from the dawn of the 20th century (Rogers, 2003), the

DOI framework has been widely embraced for studying the diffusion and adoption of IT

and Internet innovations (Andrés et al., 2010; Häggman, 2009; Ross, 2010; Valier et al.,

2008; Williams, Dwivedi, Lal, & Schwarz, 2009).

As an alternative to the DOI framework, Vega, Chiasson, and Brown (2008)

proposed the systems of innovation approach (SIA) for understanding diffusion of

complex innovation. Vega et al. described SIA as an organizational learning process

influenced by social capabilities and acquired knowledge that shaped the development,

diffusion, and use of innovations. Although cloud computing has been shaped by

complex and emergent interactions, yielding new knowledge and intermediate

innovations, the full depth and breadth of the SIA framework adds a degree of complexity

not merited by this study’s research questions. Recognizing the importance of these

influences on DOI, social factors have been engineered into the research by refining the

traditional measures related to the perceived attributes of innovation. Based on the

study’s findings, the potential for follow up SIA research may be warranted.

The traditional DOI framework has been contextualized by four main elements:

innovation, communication via channels, time duration, and social system constituency

(Häggman, 2009; Rogers, 2003). Silverstein, Samuel, and DeCarlo (2009) prefaced their

conceptualization of innovation by describing innovation in the common terms of

“introducing something new” or “coming up with the next big idea” (pp. xvii-xviii).

Rogers’s (2003) DOI research revealed that the rate of innovation adoption was

determined by five variables: perceived innovation attributes, innovation decision types,

communication channels, social system characteristics, and change agent effectiveness.

Suggesting strategies for cloud-principled computing technology innovation, Sahoo

(2009) adapted the traditional DOI framework into a triune context consisting of

customer context, industry-competitor context, and technical intricacies context. Sahoo’s

customer-value perspective context in conjunction with the technical intricacy context

closely resembled Rogers’s (2003) DOI construct for the perceived attributes of

innovation while the industry-competitor context was analogous to Rogers’s DOI social

system characteristics construct. Similarly, for the purposes of this study, the primary

DOI framework emphasis has been refined to focus on an expanded delineation of the

attributes of innovation related to cloud-computing innovation in lieu of the traditional

measures.

Understanding the perceived attributes of innovation has become a predominant

focus of DOI research (Häggman, 2009; Rouibah & Hamdy, 2009; Valier et al., 2008;

Vega et al., 2008). Rogers (2003) asserted, “The individual’s [adopter or potential

adopter] perceptions of the attributes of innovation, not the attributes as classified

objectively by experts or change agents, affect its rate of adoption” (p. 223). In his DOI

research, Rogers further distilled the perceived attributes of innovations to include

relative advantage, compatibility, complexity, trialability, and observability (pp. 15-16).

However, like other researchers, Rouibah and Hamdy’s (2009) instant messaging

adoption study and Chong, Ooi, Lin, and Raman’s (2009) collaborative commerce

(ccommerce) adoption inquiry limited the measures of innovation attributes to relative

advantage, compatibility, and complexity.

In prominent diffusion research (Moore & Benbasat, 1991; Rogers, 2003;

Venkatesh et al., 2003), relative advantage was emphasized as an essential innovation

attribute in determining the adoption of IT innovations. Often expressed in economic

value, societal significance, and other beneficial facets, the variable relative advantage

was used to measure the perceived degree of betterment attributed to an innovation in lieu

of precursor ideas (Rogers, 2003). Regularly associated with perceived usefulness,

relative advantage within the context of this study was gauged by quickly performing

tasks, improving quality, working easier, enhancing effectiveness, improving

performance, and increasing productivity.

Compatibility was defined as a measure of the perceived degree of congruence

attributed to the innovation based on the adopter’s experiences, values, and needs

(Rogers, 2003). Distinguishing compatibility as an essential innovation attribute measure

for technology acceptance, Rouibah and Hamdy (2009) further calibrated compatibility as

a measure of the adequacy of fit for the task the technology was intended to perform as

well as culture in which it will be used. Consistent with Rouibah and Hamdy’s meaning,

compatibility within the context of this study was assessed by factors including

compatibility with work, productivity, current situation, and work style.

Analyzing four determinates of c-commerce adoption: innovation attributes,

environment, information sharing culture, and organizational readiness, Chong et al.

(2009) emphasized the perceived degree of complexity as an important innovation

attribute determinant. Although commonly associated with perceived ease of use, the

variable complexity was defined as an assessment of the perceived degree of difficulty or

simplicity attributed to understanding and using the innovation (Chong et al., 2009;

Conrad, 2010; Rogers, 2003). Chong et al. and Rogers’s (2003) meaning of complexity,

which gauges factors such as cumbersome to use, understandability, performs as desired,

and easy to learn, was deemed consistent with the purpose of this study.

Trialability was defined as a measure of the perceived ability for use inspection

attributed to the innovation (Conrad, 2010) while observability is the perceived degree of

communicability or visibility to others attributed to the innovation’s use, features, or

benefits (Rogers, 2003). Although Rouibah and Hamdy (2009) and Chong et al. (2009)

eliminated these measures from their innovation adoption research, for this study both

trialability and observability were considered important attributes of innovation for cloud

computing based on the obstacles associated with cloud computing revealed in the

literature review.

In the context of this study, Rogers’s (2003) meanings of trialability and

observability were applied. Rogers defined observability as a measurement used to

examine other firms’ use, other internal uses, the degree visible, and the ease of viewing

other uses of the innovation. Additionally, Rogers defined trialability as a gauge of the

opportunity for experimentation, knowing where to try, the ability to try, and the

empowerment to try or experiment adequately.

Building upon the attributes of innovation research, Moore and Benbasat (1991)

developed a generalizable survey instrument consisting of 25 scale items to measure eight

attributes of IT innovations. The eight attributes consisted of a hybrid of Rogers’s classic

five plus three additional attributes. Moore and Benbasat’s additional attributes included

social variables: image, voluntariness, and results demonstrable. Image was described as

the perceived degree of enhancement of one’s social system status or prestige (Moore &

Benbasat, 1991). Voluntariness was depicted as the perceived degree of self-determined

sovereign choice to use an innovation (Moore & Benbasat, 1991).

Extending and superseding the theory of reasoned action (TRA), researchers’ use

of TAM has dominated IT and Internet inquiry related to ease of use and usefulness

acceptance measures (Azadegan & Teich, 2010; Davis, 1989; Venkatesh et al., 2003).

Moore and Benbasat embraced the influence of Davis’s (1989) TAM research by

integrating the construct perceived usefulness into the classic innovation attribute relative

advantage and superseding the attribute complexity by the construct perceived ease of use

(Moore & Benbasat, 1991; Rogers, 2003). Al-Hajri and Tatnall (2008) highlighted the

variable ease of use as a focal measure driving Internet banking adoption. Moore and

Benbasat (1991) further delineated the classic attribute observability by its distinct

attributes of communicability and visibility resulting in the refined attribute measures of

results demonstrable and visibility.

Researchers applying DOI theory typically have examined the rate of innovation

adoption based on five variables. Researchers employing TAM primarily have correlated

the constructs of perceived usefulness and perceived ease of use with the acceptance or

rejection of technology, in particular IT. DOI researchers were concerned with

innovations, technology based and otherwise, whereas TAM researchers are singularly

interested in technologies. Davis (1989) defined perceived usefulness as the extent to

which a person’s belief that using a particular IT system enhances his or her job

performance. Likewise, Davis defined perceived ease-of-use as the degree to which a

person believes that using a particular system would be effortless or free of difficulty.

TAM has been used extensively in research to examine user acceptance of various IT

innovations (Al-Hajri & Tatnall, 2008; Gounaris & Koritos, 2008; Williams et al., 2009).

Expansive IT acceptability research has led to the proliferation of technology

acceptance and adoption models. Venkatesh et al. (2003) evaluated eight prominent

models related to an individual’s intentions to use IT culminating in the unified theory of

acceptance and use of technology (UTAUT). Venkatesh et al.’s UTAUT model consisted

of four core variable determinates of technology acceptance and up to four moderating

variables. Gounaris and Koritos (2008) emphasized Moore and Benbasat’s (1991)

usability attributes while also incorporating measurement factors related to the social and

psychological aspects of the adoption process. While recounting the parsimonious

attraction of the TAM framework, Straub and Burton-Jones (2007) suggested that

UTAUT and other TAM adaptations are far from parsimonious having a bias for common

methods variance. Straub and Burton-Jones called for comprehensive metaanalysis of

TAM’s independent and dependent variables, their relationships, and moderating factors

reconstituting TAM into its essential parsimonious set of variables and critical

antecedents.

Additionally, Compeau, Meister, and Higgins (2007) extended Moore and

Benbasat’s (1991) theoretical framework further transforming the variables compatibility

and results demonstrability. Compeau et al. extrapolated the variable compatibility into

three constructs: compatibility with prior experience, preferred work style, and values.

Expressing concern related to the potential inability to distinguish between Moore and

Benbasat’s constructs for visibility and results demonstrable, Compeau et al. added clarity

to visibility by emphasizing visibility of others’ use and transforming the variable results

demonstrability into two constructs: communicability and measurability.

As previously stated, Rogers’s (2003) framework defining the variable

compatibility and observability were preferred in lieu of Moore and Benbasat’s (1991)

and Compeau et al.’s (2007) framework related to these constructs. However, Moore and

Benbasat’s constructs for results demonstrable and voluntariness, which have a social

context, were considered appropriate based on the marketing hype and policy mandates

exposed in the literature review. For the purposes of this study, the construct for results

demonstrable was defined to assess factors such as difficulty in telling others, sharing

consequences, apparent outcomes, and explaining the rationale for use, while the

construct voluntariness was defined to gauge indicators such as management’s use

expectation, adopter’s choice, optional use, compulsory use, and preferences.

Acceptability research typically has been forward looking, investigating the ideal

attributes of an innovation, to position the innovation for greater acceptance and,

therefore, a more rapid rate of adoption (Rogers, 2003; Valier et al., 2008). As a point of

reference, typical DOI research is intrinsically postdictive correlating present perceived

attributes of innovation with the prior adoption rate of the innovation (Valier et al., 2008).

In their diffusion of the Internet research analyzing longitudinal data collected over the

period of 1990 through 2004 in 214 countries, Andrés et al. (2010) examined actual

innovation adoption versus the potential for adoption signaling a traditional postdictive

approach. Acknowledging that the number of producers engaged in offering an

innovation has been used as a measurement of its diffusion, Andrés et al. measured

diffusion based on user adoption that was more consistent with Rogers’s (2003) DOI

approach and the theoretical framework of this study.

Recognizing the postdictive nature of DOI research, Rogers (2003) advocated a

prediction DOI research design that measured the attributes of innovation at an

innovation’s prediffusion stage (time1), which is a time prior to, or concurrent with, an

individual’s innovation adoption decision, to predict its rate of adoption during future

diffusion (time2). Rogers favored this predictive research design strategy in lieu of (a)

extrapolating the rate of adoption from a prior similar innovation into the future or (b)

measuring the potential adopters’ perceived attributes of a hypothetical innovation to

predict its impending rate of adoption. Alternatively, Häggman (2009) demonstrated

using the DOI framework as means of inductive inquiry related to adoption and diffusion

of technology.

Compeau et al. (2007) deepened the conceptualization of predictive DOI research

focusing beyond the direct effects of perceived attributes of innovation. Compeau et al.

investigated the ways in which antecedents to innovation adoption operated to influence

the behavior of innovation adoption and use. Evolving Rogers’s (2003) perceived

attributes of innovation through the lens of Moore and Benbasat’s (1991) acceptance

constructs, Compeau et al. developed a more refined set of variables to measure

innovation adoption that undergirded their extended theoretical DOI model of indirect

antecedent influences of innovation adoption. Compeau et al.’s extended predictive DOI

model offered a greater scope than required for this study; their research further

substantiated the direct effects of the independent variables examined in this doctoral

research.

Conceived in Rogers’s (2003) predictive DOI research assumptions, Valier et al.

(2008) developed a theoretical model to test the predictive ability of DOI theory. Valier

et al.’s predictive DOI model was used to examine, during the prediffusion stage, the

correlation between seven perceived attributes of an innovation and the potential

adopter’s intent to use the innovation. The activities leading to the early adoption of an

innovation, which signal the beginning of diffusion, encompass the prediffusion stages of

the innovation-development process (Rogers, 2003). However, Valier et al.’s research

limited prediffusion activities through the research and development stages, thereby

excluding the commercialization stage. The conceptual series of the

innovationdevelopment stages depicting the predecessor role of the prediffusion stages is

illustrated in Figure 1.

Problem

Identification

Research

Diffusion

& Adoption

evelopment

Commer-

cialization

onse

uence

Postdictive:

Traditional Tracer Studies

Traditional DOI Studies

Predictive:

Prediffuson DOI Studies

Figure 1. Theoretical sequence of activities in the innovation-development process.

Adapted from “Diffusion of Innovation, 5th Edition” by Everett M. Rogers, 2003, The

Free Press, a Division of Simon & Schuster, Inc. Reprinted with permission of the

Constructed on the innovation research of Rogers (2003), Moore and Benbasat

(1991), and Gounaris and Koritos (2008), Valier et al.’s (2008) predictive DOI model

included the seven independent variables: relative advantage, compatibility, complexity,

results demonstrable, trialability, observability, and voluntariness. The dependent

variable in Valier et al.’s model consisted of intent to use, which was the outcome of the

decision stage, the third of five stages, in the innovation-decision process (Rogers, 2003,

p. 169). The innovation-decision process is illustrated in Figure 2.

- - - - - - - - - - Communications Channels - - - - - - - - - -

4.Trialability

5.Observability

Figure 2. Theoretical continuum of decision-making activities in the innovation-decision

process. Adapted from “Diffusion of Innovation, 5th Edition” by Everett M. Rogers,

Knowledge

Prior

Conditions

II.

Persuasion

III.

Decision

IV.

mplement

Confirmation

.Prior practice

.Felt needs

.Innovativeness

.Social Norms

Decision Maker

Attributes

.Socioeconomic

.Personalities

.Communicable

erceived

ttribute

of Innovation

.Relative

Advantage

.Compatibility

.Complexity

. Adoption

. Rejection

Use

Later Adoption

Discontinuance

Non-Use

Valier et al. (2008) successfully tested their predictive DOI model by measuring,

during the prediffusion stage, the perceived attributes of open source software

innovations and Linux User Groups World Wide members’ use intentions. The

theoretical framework applied to this doctoral study was adopted from Valier et al.’s

predictive DOI model based on Rogers’s (2003) four elements of DOI. In this instance,

the four elements of DOI were applied as follows:

1. The innovation was emerging cloud-computing technologies.

2. The channels of communications were both formal and informal.

3. The timeframe was prediffusion through the commercialization stage.

4. The social system included the Arizona small business sector.

Gounaris and Koritos (2008) extended the innovation-attributed framework

combining measures of usability associated with TAM, attributes of innovation

characteristic of DOI, and social and psychological measures. Using a logistic

regression, Gounaris and Koritos analyzed the relationship of the various measurement

factors and the predictability of Internet banking adoption. Gounaris and Koritos’s

research closely resembles the predictive DOI model used by Valier et al. (2008).

Additionally, this study’s independent and dependent variables were adopted from

Valier et al.’s (2008) predictive DOI model. The independent and dependent variables

were examined to test empirically the hypothesis derived from this study’s research

questions. The independent variables were defined to assess the perceived attributes of

cloud-computing innovation, which included relative advantage, compatibility,

complexity, results demonstrable, trialability, observability, and voluntariness. Consistent

with the predictive DOI model, these independent variable measurements reflected the

potential adopter’s perceptions versus those of industry experts or DOI researchers. The

dependent variable, intent to use, was defined to assess small business leaders’ propensity

to adopt cloud-computing innovation. Operationalized definitions of the study variables

are stipulated in the Data Collection segment in Section 2.

Although the innovation-development process featured in Figure 1 depicted a

series of sequential stages, Rogers (2003) affirmed that the timing, occurrence, and

iteration of stages were not uniform for all innovations because of reinvention and other

factors influencing the process. Cloud-computing innovations have fluctuated between

the development processes with standardization of services yet fully defined for

commercialization. Cloud computing was envisioned as utility-like commercialization

demonstrating rapid elasticity of measured services that were not generally available to

all potential adopters. The theoretical framework for this DOI inquiry is illustrated in

Figure 3 depicting the research, development, and commercialization stages of the

innovation-development process.

Figure 3. Theoretical predictive DOI framework. Framed in context of the

innovationdevelopment and innovation-decision processes, the perceived innovation

attributes were related via the hypotheses to the potential adopter’s use intention. The

innovationdevelopment process and innovation-decision process were adapted from

“Diffusion of Innovation, 5th Edition” by Everett M. Rogers, 2003, The Free Press,

by Free Press, a Division of Simon & Schuster, Inc. Reprinted with permission of the

Definition of Terms

This study was framed on concepts and terminology centric to Internet and

cloudcomputing technology that may be unfamiliar to the reader. A brief description was

provided explaining the meaning for each key technology term.

Broad network access: Extensive connectivity and usability via standard

mechanisms supporting heterogeneous thin or thick client platforms (Mell & Grance,

2010).

Cloud computing: Large, scalable, on-demand, rapidly provisioned, IT assets and

capabilities ecosystem connected with and available through the Internet including

networks, servers, storage, applications, and services (Cheng, 2010; Mell & Grance,

2010; Smith, 2009).

Community cloud: A cloud infrastructure collectively supporting organizations

that have a shared affinity, concerns, or purpose (Mell & Grance, 2010).

Competitive advantage: A position of strategic benefit acquired by a firm in

relation to its competitors (Porter, 1998).

Diffusion: The process of innovation communication via channels to members of

social systems over time (Rogers, 2003).

Information technology: The management of computer-based information systems

relating to software applications and computer hardware used to convert, store, protect,

process, retrieve with security, or transmit any information (Information Technology

Association of America, 2009).

Infrastructure as a Service (IaaS): The provision of managed processing, storage,

networks, and other fundamental computing resources for consumer deployment of

software including operating systems and application software via the Internet (Mell &

Grance, 2010).

Innovation: The embodiment, combination, or synthesis of knowledge or an idea

related to relevant products, processes, services, or other objects that are perceived by its

adopter as novel and potentially value adding (Crespell & Hansen, 2008; Rogers, 2003).

Innovation-development process: The innovation life-cycle continuum of activities

and decisions that transpires commencing with the perception of a problem need through

research, innovation development, commercialization, diffusion and adoption, and

consequence (Rogers, 2003).

Internet: The global information system that (a) is logically linked by a globally

unique address space based on the Internet Protocol (IP) or its subsequent extensions; (b)

can support communications using the Transmission Control Protocol/Internet Protocol

(TCP/IP) suite or its subsequent extensions, or other IP-compatible protocols; and (c)

provides, uses or makes accessible, either publicly or privately, high-level services

layered on the communications and related infrastructure (Federal Networking Council,

1985).

Hybrid cloud: A cloud infrastructure comprising two or more clouds (private,

community, or public) that is bound together by standardized or proprietary technology

that enables data and application portability (Mell & Grance, 2010).

Measured Service: The transparent provisioning, metering, and accounting of an

abstraction of computing resources in accordance with a service level commitment (Mell

& Grance, 2010).

On-demand self-service: A consumer’s unilaterally provisioned computing

capabilities as needed without requiring service provider human interaction (Mell &

Grance, 2010).

Platform as a Service (PaaS): The computing capability deployed onto the cloud

infrastructure consisting of programming environments, virtualization, layered interfaces,

and other development tools supplied and maintained by the service provider enabling

consumers to use created or acquired applications via the Internet (Mell & Grance, 2010).

Prediffusion: The activities in the innovation-development process leading to the

commencement of diffusion, which is signaled by the first adoption of the innovation

(Rogers, 2003; Valier et al., 2008).

Private cloud: A cloud infrastructure operated exclusively for a sole organization

(Mell & Grance, 2010).

Public cloud: A cloud infrastructure commercially available to the general public

or a large industry group (Mell & Grance, 2010).

Rapid elasticity: The seemingly infinite dynamic and immediate provisioning of

computing resources that scales (up or down) to instant consumer demand (Mell &

Grance, 2010).

Resource pooling: The autonomous dynamic multi-consumer sharing of

computing assets or components (Mell & Grance, 2010).

Software as a Service (SaaS): The provision of supplier managed application

programs to consumers hosted on a cloud infrastructure (Katzan & Dowling, 2010; Mell

& Grance, 2010).

Service Oriented Architecture (SOA): An open standard based on loosely coupled

components and a functional architecture that allows the integration and interoperability

of different applications by different members of an IT ecosystem (Sahoo, 2009; Vouk;

2008).

Small business: A size classification of business entities organized for profit. In

the United States, the SBA established small business size standards on an industry-

byindustry basis as a function of annual revenues or number of employees, which

typically include firms employing fewer than 500 employees and generating revenues

below $50 million (SBA, 2007).

Sustainability: To establish and maintain economic, environmental, societal

viability for the present as well as the future (Senge, Kruschwitz, Laur, Schley, & Smith,

2008).

Assumptions, Limitations, and Delimitations

The assumptions, limitations, and delimitations related to this research study were

disclosed to enhance reviewer comprehension as well as to inform further research about

this study’s methods, conclusions, and findings. The assumptions were defined to

recognize unconfirmed facts considered true that may beget potential risks. The

limitations were revealed to highlight known areas of restrictions potentially exposing

areas of study deficiency. The delimitations were defined to delineate the study’s scope

and boundaries.

Assumptions

The first assumption warranting disclosure posited that the IT agility, efficiency,

and economic benefits attributed to large organization’s early adopter use of emerging

cloud-computing innovation similarly would be experienced by small businesses, thus

resulting in their increased economic viability and sustainability. Process agility,

computing scalability, outsourced technology competency, economies of scale, and

operationalized costs were identified in the literature review as rationale for large

company and public agency early adoption of the yet fully commercialized

cloudcomputing innovations. Although large-firm practices are not necessarily

transferable to small businesses (Galán, Monje, & Zúñiga-Vicente, 2009; Tan, Fischer,

Mitchell, & Phan, 2009), the large enterprise cloud-computing innovation benefits were

anticipated applicable.

Another assumption was that the individuals from small businesses participating

in this study had sufficient understanding of emerging cloud-computing technology to

record their perceptions regarding its innovation attributes. Although cloud-computing

technology may not be fully evolved through the research and development and

commercialization stages of the innovation-development process continuum,

cloudcomputing innovation has been sufficiently conceptualized and communicated via

formal and informal channels. As a result, small business participants were believed

informed sufficiently to conceptualize perceptions about the cloud-computing innovation

attributes.

To add clarity, cloud-computing innovation terminology was specified in the study survey

instrument to mitigate risks associated with this assumption.

A related assumption was that of participant self-reporting bias, which is

influenced by societal and cultural norms expressed via personal feelings and attitudes.

Moreover, self-reporting bias occurs when a participant’s experience, self-perception, and

work environment influence their survey responses (Han & Anantatmula, 2007).

Participant self-reporting bias was mitigated by soliciting honest and objective participant

survey responses.

Limitations

A common limitation of DOI research has been a pro-innovation bias. This

phenomenon in DOI research was conceptualized as a predetermination that the subject

innovation should be adopted and spread rapidly throughout all members of the targeted

social system without consideration of re-innovation or rejection (Rogers, 2003). A

proinnovation bias has limited researchers’ vision to overlook or underemphasize

innovation rejection, discontinuance, or re-invention as well as to avoid antidiffusion

inquiry.

The rationale for the selection of cloud-computing technology as the innovation of

interest for this DOI research mitigated the pro-innovation bias. Specifically, the

preliminary research conducted to uncover and understand the problem as stipulated for

this research study evidenced the meritorious IT agility and economic value of

cloudcomputing innovation while recognizing its embryonic development. The re-

inventive nature related to certain aspects of cloud-computing technology was a driving

factor in the election of its prediffusion study emphasis.

The prediffusion emphasis, which included the innovation-development stages

leading to the innovation adoption decision, of this study was narrowed by its predictive

focus. Consideration of the adoption rate, adoption consequences, or other influences

associated with innovation diffusion was excluded because of this study’s predictive

emphasis. These additional innovation diffusion considerations may delineate the scope

of further inquiry beyond this research study. Additionally, Valier et al. (2008)

recognized the predictive methodological approach as potentially limiting the

appropriateness of causal relationship analysis.

Another potential limitation of this study was related to the practical limitations in

assembling a list of the entire Arizona small business sector as a whole. The population

sampling frame elected was convenience sampling (Creswell, 2009) based on the entire

small business e-mail list maintained by the ACA. As a result, generalization of the

study’s findings was limited potentially to the study population and was not implied

beyond. The potential convenience sample limitations were mitigated to a degree by the

anticipated increased sample size.

Delimitations

The problem statement, theoretical framework, and research design have been

instrumental in determining the scope and bounds related to this research study. The

delimitation discussed in this study proposal included four factors: innovation election,

innovation measurement, population sample selection, and data collection and analysis.

The boundaries associated with these delimitation factors were used as a guide to identify

what was in and out of the project scope.

After a preliminary scan of emerging technology innovations, cloud-computing

innovation’s meritorious business computing paradigm was revealed as a potential means

for greater small business economic contribution. Although this insight potentially

reinforced a pro-innovation bias, the potential economic and societal benefits warranted

further investigation to understand small business leaders’ propensity to adopt

cloudcomputing innovation relative to its perceived innovation attributes. This DOI

research project was delimited by the specific innovation of interest, which was cloud-

computing technology.

Encompassing a wide range of exploratory possibilities spanning the six stages of

the innovation-development process continuum, the five stages of the innovation-decision

process spectrum, and the five variable determinants of the rate of adoption, DOI research

has been extensively embraced (Rogers, 2003). This predictive DOI research project was

delimited by the prediffusion stages in the innovation-development process continuum,

the decision stage in the innovation-decision process spectrum, and the perceived

attributes of innovation in the determination of the rate of adoption arena.

The DOI social system of interest was delimited to the small business sector.

More specifically, Arizona small businesses delimited this study’s subject scope. Based

on the research design, data collection was performed during a 10-day period using a

cross-sectional online survey. Additionally, quantitative statistical methods delimited the

data analysis conducted. These study delimitations, in whole, composed this distinctively

unique research study.

Significance of the Study

This study’s significance was demonstrated by reducing the gap identified in

extant research concerning small businesses’ penchant to adopt cloud-computing

innovation and by recognizing the potential implications for positive social change. The

research gap reduction was prefaced by a recap of extant research deficiency. The

implications for social change extended beyond small business sustainability affecting

economic, societal, and environmental benefits.

Reduction of Gaps

The value and volatility of small businesses in the U.S. economy and free

enterprise capitalism has permeated the literature. The importance of IT and innovation

adoption for competitive advantage in support of small businesses’ sustainability was

broadly evidenced throughout scholarly research. However, the review of the literature

revealed a deficiency in understanding about the propensity of small businesses to adopt

emerging cloud-computing innovation, relative to the relationship of perceived

cloudcomputing innovation attributes and adoption intent. This research has strengthened

the potential for increased small business economic viability and sustainability via the

increased understanding about the relationship of small businesses’ perceived

cloudcomputing innovation attributes and their intent to use this emerging innovation.

Building on the framework of prior DOI research, this research was conducted to

address the literature gap by examining small businesses’ propensity to adopt

cloudcomputing innovation. More specifically, this study’s findings revealed the

correlation of small businesses’ prediffusion perception of cloud-computing innovation

attributes in conjunction with the intent to use cloud-computing technology. A limited

number of studies have been conducted based on the prediffusion stage of innovation

adoption while the preponderance of DOI research has been conducted during and post

innovation adoption. Filling this void in the literature provided a steppingstone for

continued research linking cloud-computing technology to small business competitive

advantage and sustainability.

Implications for Social Change

Business sustainability was defined as a triune conceptualization consisting of (a)

making present choices that do not compromise future choices, (b) enhancing the

economic well-being, and (c) respecting the ecological and societal carrying capacity

(Senge et al., 2008; Will, 2008). A better understanding of small businesses’ potential use

of cloud-computing innovation holds the possibility of reducing small business failures,

increasing their economic contributions, and promoting sustainability. Furthermore, this

study’s findings and added understanding was anticipated to transcend the benefits of

business sustainability by potentially guiding policy practices for elevating employee job

skills, improving private and public sector collaboration, and increasing the U.S. gross

domestic product and U.S. global innovative leadership (Estrin, 2009).

Small businesses may use this study’s findings to improve capitalization of IT

innovations for sustainability. These findings were anticipated useful for projecting

employee development proficiencies in preparation for deploying cloud-computing

innovations. Local communities were expected to benefit from these findings by

understanding the importance for educating the nascent workforce related to developing

cloud-computing technology skills. Regulatory and public agencies were anticipated to

develop better insights related to assisting small business employers with needed

technology innovation collaboration support to facilitate academic and public sector

technology commercialization (Blau, 2009; Renski, 2009; Rosser & Taylor, 2008).

Additionally, rapid diffusion of cloud-computing innovation has the potential for

significant positive environmental impact via the reduction of overall energy

consumption and electronic equipment disposal waste shifting the burden from nature

(Senge et al., 2008, pp. 199-200).

A Review of the Professional and Academic Literature

The literature reviewed for this research project was organized into five topical

areas. The first topic area was small business perspectives, which identified current

trends and information relative to small business statistics, economic contributions,

global impact, and closures. The second topic area was small business innovation policy,

which surfaced trends in small business innovation collaboration, governmental

intervention, and academic influences. The third topic area was small business

innovation essentials, which revealed the criticality of business innovation, the strategic

role of innovation, and trends in small business innovations. The fourth topic area was

small business IT innovation, which focused on small business IT and Internet innovation

practices. The fifth topic area was cloud-computing ecosystems, which presented a

conceptualization of this rapidly evolving technology emphasis.

Although printed book publications were included in the literature review, the

main literature search was conducted via electronic databases, primarily EBSCOHost’s

Computers and Applied Sciences Complete and Business Source Complete/Premier,

ProQuest’s ABI/INFORM Complete and Business Dissertations and Theses, and Emerald

Management Journals. Keywords used for the electronic database search of the topic

domains included x business sustainability, entrepreneur, small business, U. S. economy

for small business perspectives; x collaborative commerce, diffusion of innovation,

economic policy, and technology transfer policy for small business innovation policy; x

competitive advantage, innovation, technology invention, pre-adoption, and product

development for small business innovation essentials; x e-business, e-commerce, Internet,

and IT strategy for small business IT innovation; and x cloud computing, grid computing,

service computing, utility computing, and virtualization for cloud-computing ecosystems.

A total of 152 resources were found relevant, eight of them were books, two of them were

dissertations, seven of them were government publications, and two of them were online

references. The remainder was peer-reviewed journal articles. A map of the literature

reviewed is presented in Appendix C to communicate the literature’s basis for

establishing ongoing inquiry and research.

An understanding of the importance of small businesses in the United States and

to the global economy was developed based upon the literature review. This

understanding included factors affecting the efficacy of small businesses as well as

enabling influences that foster innovation in small business. Additionally, current

motivations and strategies regarding the innovation practices of small business were

explored. Last, the rapidly evolving presence of cloud computing was examined as a

potential ecosystem for novel small business innovation to facilitate sustainable

competitive advantage.

Small Business Perspectives

The literature review concerning small business perspectives was highlighted to

build an understanding of current trends and information relative to small business mass

and economic contributions, small business global impact, and small business closure or

discontinuance volatility. The SBA has recognized the critical importance of small

business to the United States’ economic strength and its role in the global marketplace.

Since its establishment in 1953, the SBA (2009) has served and protected the interests of

small business concerns to preserve free competitive enterprise. Based on 2008 data, the

SBA (2009) reported that the more than 27 million small businesses in the United States

represented 99.7% of all employer firms, generated 60 to 80% of net new jobs annually,

and created more than one half of the nonfarm private gross domestic product.

The U.S. statistics were similar to those reported relative to comparable free

enterprise markets. In Australia, small businesses accounted for 96.7% of total business

numbers employing 70% of the total workforce (Quaddus & Hofmeyer, 2007; Scupola,

2009). In Europe, small businesses numbered 23 million and account for 99% of all

European businesses (Blau, 2009). The similarity of these statistics depicted small

businesses’ global economic significance.

Defining exactly what constituted a small business was challenging. Sovereign

governing bodies throughout the world have established unique criteria for designating a

firm’s size. The U.S. federal government has stipulated a business’ size as a function of

annual revenues or number of employees based on the North American Industry

Classification (NAIC) system (Small Business Act, 1979). The maximum qualifying

annual revenue varies from $750 thousand to $50 million and the maximum number of

employees ranges from 100 to 1,500 depending on NAIC designation (SBA, 2007).

Although in the United States small businesses typically included firms employing fewer

than 500 employees and generating revenues of $50 million or less, small businesses

were reported to comprise nearly all U.S. employers, employ the majority of the private

sector workforce, and create the greater part of the gross domestic product (SBA, 2009).

The executive arm of the European Union defined small and medium size

enterprises (SMEs), which were comparable to U.S. small businesses, as independent

companies with fewer than 250 employees (Blau, 2009). In Australia, small businesses

were designated as firms employing 200 or fewer workers (Scupola, 2009). However, in

New Zealand SMEs were considered firms with fewer than 20 employees (Al-Qirim,

2007). These various definitions created ambiguity in identifying small businesses with

consistency throughout the literature. Universally, SMEs have been characterized as

catalytic economic agents with relatively limited resources and broad diversity (Forsman,

2008).

To counteract the decline in the United States’ leadership role in technology due to

fewer workers and entrepreneurs entering careers in competitive science, technology,

engineering, and mathematics (Rosser & Taylor, 2008), strengthening and expanding

small businesses’ innovation capabilities has been defined as a top priority (Galston,

2010; Heffes, 2009; SBA, 2009). In 2008, the European Commission approved the EU

Small Business Act, which was designed to help small businesses and entrepreneurial

startups to develop innovative capacity as a means of closing its research gap with the

United States (Blau, 2009). Barba-Sánchez and Martínez-Ruiz (2009) emphasized

European SMEs’ contribution to regional employment and social-economic development.

Li and Mitchell (2009) recognized the competitive dynamics of the Chinese knowledge

worker spillover as a model to stimulate radical small business innovation in transitional

economies. In comparison, small businesses in the United Kingdom’s more developed

economy have been predisposed to focus more on leveraging return on investments, thus

favor incremental versus radical innovations (Oke et al., 2007). Additionally, Uddin’s

(2006) study of innovations diffusion in Bangladesh offers an important consideration for

incremental change leading to sustained small business development globally. These

global considerations for small business innovation and technology leadership

demonstrated the economic value of small business innovations.

Kampschroeder, Ludwig, Murray, and Padmanabhan (2008) exposed the adverse

rippling economic fallout of failed small businesses. Faced with large-company and

global competition, small businesses have experienced high discontinuance rates: 76% of

new firms remain open after 2 years, 47% after 4 years, and 38% after 6 years (Liao et

al., 2008; SBA, 2009). Likewise, Tan et al. (2009) reported that somewhere between

50% and 80% of small businesses fail.

At the state level, year-to-date third quarter 2008 Arizona small business

discontinuances outpaced new ventures by 13.75% and small business contractions

exceeded expansions by 44.7% (SBA Office of Advocacy, 2009). Nonfarm Arizona

small business employers in 2006 made up approximately 1.8% of U.S. small business

employers (SBA, 2009). During the 2006 timeframe, Arizona small business employers

numbered approximately 107,500 firms, accounted for 97.4% of the state's employers,

and employed 48.8% of the state’s private-sector workforce (SBA, Office of Advocacy,

2009). More recently, the U.S. Census Bureau (2008) reported that Arizona small

businesses numbered approximately 106,800 firms, reflecting a decline in small business

economic vitality.

However, equating small business closure with business failure was misleading.

Bates (2005) reported that according to U.S. Bureau of the Census survey data,

approximately 37% of Year 6 firm closures were deemed successful at the time that the

decision was made to cease operations. Understanding the rationale for small business

discontinuance was an important consideration in their economic contributions. A

resource-based view of nascent entrepreneurs revealed that financial resources and

technology knowledge were salient factors in business discontinuance (Liao et al., 2008).

However, the literature review exposed a gap in knowledge in relation to cloudcomputing

innovation and small business adoption for economic sustainability.

Small Business Innovation Policy

The literature review in connection with small business innovation policy

explored the rationale for governmental intervention, investigated academic knowledge

transfer influences, and highlighted business collaborative trends in small business

innovation. Recognizing the commercialization value of small businesses for public

sector research, U.S. federal policy established the Small Business Innovation Research

(SBIR) and Small Business Technology Transfer (STTR) programs. SBIR and STTR

programs collaboratively team private sector expertise via small business entrepreneurs

and innovative university research with public sector funding to maximize return on

investment for research, foster innovation leadership, and stimulate economic

development (Ford, Shino, Sander, & Hardin, 2008).

Similarly, state government and local municipal public policies such as ACA’s

Angel Investor and Enterprise Zone programs promoted local and regional economic

development via collaborative public assisted research and small business

commercialization (ACA, 2011). ACA is a newly established quasi-public state authority

governed by a board of directors comprising public and private sector members

envisioned to foster an operating environment for strategic collaboration and economic

growth. Emphasizing small business and entrepreneurship, ACA’s strategies were aimed

at diversifying the overall state economy by incubating and bringing in new businesses.

Often these policies have resulted in agglomeration of research and small business

incubation clusters further leveraging economies of scale and network effects (Maine,

Shapiro, & Vining, 2010).

Although government intervention aimed at information and communication

technologies (ICT) in rural economies was designed to foster regional small business

sustainability and growth, government-provided Internet technology to rural communities

appeared to stifle regional entrepreneurship as a result of greater competitive access to

rural communities (Cumming & Johan, 2010). Analyzing the multiscalar innovation and

development research-center frameworks in North America, Clark (2010) recommended

horizontal and vertical coordination policy within national innovation systems evidencing

the overall value attributed to innovation commercialization as an economic policy

intervention. Government policy intervention was intended to serve as a catalyst and

resource for small business innovations (Michael & Pearce, 2009).

Likewise, university-based research centers have fostered commercialization of

innovation via small businesses to leverage the economic value of their research and to

further academic pursuits. For example, the Center for Advanced Technology and

Innovation pioneered a unique technology-led economic development strategy in

Southeastern Wisconsin to foster rural small business innovation transfer (Wagner, 2008).

Similar research centers designed for commercialization of technology innovation have

emerged in other global regions. For example, Finland has established its global

competitiveness by retaining economic value of homeland production through academic

research technology transfers bringing together technology, design and small business to

develop innovative products (Cervi, 2008).

An important motive for collaborative small business innovation was to create

economies of scale for market share benefit (Allen & Stearns, 2009; Carree & Thurik,

2008). Allen and Stearns (2009) highlighted small businesses in rural communities

successfully competing with urban centers via regional innovation and technology

entrepreneurship collaboration. Similarly, the Organization for Economic Co-operation

and Development (OECD) has fostered sustainable economic growth, advanced

employment, and promoted individual living standards by lobbying democratic

governments and market economies to develop economic cooperatives that innovatively

and competitively cultivate small business development (Carree & Thurik, 2008). In

recent years, the European countries have developed collaborative economic policies

based on economic regional development models to stimulate the creation and expansion

of small and medium-sized companies in synergistic industries (Barba-Sánchez &

Martínez-Ruiz, 2009). Another motivation for innovative collaboration was the

byproduct of operational efficiencies and cost savings generated through small business

business-to-business (B2B) trading exchanges and information sharing (Quaddus &

Hofmeyer, 2007).

Throughout the Cold War era of the mid-to-late 20th century, private and public

research centers in the United States collaboratively launched a technological revolution

that has reshaped the way people live, work, and play, and have built the most advanced

communications, computing, and scientific technology in the world (Estrin, 2009).

However, changing regulatory and market driven policies have reconstituted the once

mammoth and broad R&D center into more narrowly focused, cost-conscious innovative

strategies, thus eroding the innovation ecosystem. To capitalize on the economic value

associated with their innovation and market agility, small businesses have benefited from

governing policies compensating for their capital limitations (Etro, 2009), thus decreasing

their susceptibility to discontinuance (Liao et al., 2008). Further research was deemed

necessary to understand better small business collaboration relative to cloud-computing

adoption and sustained economic advantage, whether it is B2B, governmental, or

academic.

Small Business Innovation Essentials

The literature was reviewed related to small business innovation essentials

exploring the critical aspect of business innovation for survival, the strategic role of

innovation in small businesses, and trends in small business innovation. Commonly

understood in today’s global economy and competitive environment, a firm must

innovate or it will perish (Silverstein et al., 2009). This phenomenon has been

conceptualized by the theory of economic development as the process of creative

destruction, suggesting that innovative business activities are attributed to economic

growth and sustainability (Gúzman-Cuevas, Cáceres-Carrasco, & Soriano, 2009).

Sharpening differences between firms versus reducing them, IT innovation has been

shown to complement business management ability to achieve rapid process innovations

(McAfee & Brynjolfsson, 2008).

The vitality of American culture and free enterprise has been sustained by the

ideals of innovation, exploration, freedom, and renewal (Estrin, 2009). In their extensive

review of the literature related to small business research, Tan et al. (2009, p. 234)

concluded that small businesses were at the center of action in technology and

innovation. The various aspects of innovation have emphasized product, process, market,

supply chain, and industrial organization (Porter, 1998). Understanding the ability to

measure innovation as well as the enablers of innovation has been paramount in recent

research (Braganza et al., 2009; Mangelsdorf, 2009; Rogers, 2003). This understanding

holds true for large companies and small businesses alike (Dibrell et al., 2008) even

though Welsh and White (1981) recognized that a small business is a distinct innovation

agent and not merely a little big business.

Although the majority of family owned firms were small businesses, Ibrahim et al.

(2008) reported that 34% of the companies listed on the Standard and Poor's 500 were

family operated businesses that typically outperformed nonfamily firms. Ibrahim et al.’s

research revealed that family owned small businesses typically developed unique

strategies, which were linked to attributes conducive to innovation deriving competitive

advantage. Although small businesses typically were nimbler, more customer-oriented

and quality focused, and more active in the community (Ibrahim et al., 2008), Galán et al.

(2009) noted that small businesses experienced more trouble than larger firms in enduring

under hostile environmental conditions, suggesting other factors constrain their

sustainability.

Countering the intensified small business challenges attributed to recent complex

and turbulent economic assaults (SBA, 2009), Ribeiro-Soriano and Urbano (2009)

advocated leveraging the benefit of collaborative entrepreneurship. Ribeiro-Soriano and

Urbano’s integrated innovation and knowledge management approach was framed by

three dimensions: a shared collaborative projects strategy, adaptability via collaborative

environmental structure, and a management philosophy consisting of shared values and

trust. Galán et al. (2009) further identified how the constraints hindering small business

during difficult times were transformed or leveraged into strategic innovations for

competitive advantages. Likewise, Will (2008) advocated a corporate foresight strategy

to transfer methods of predictive and technology analysis to the small business context

for sustainable development, small business management, and strategic innovation.

These strategies and others portrayed innovation as a core small business objective.

However, Forsman (2008) claimed that there were many unanswered questions

regarding small business development projects relative to improved business

performance. Forsman further posited that the small business innovation projects seemed

to exhibit several interrelated dimensions; success propelled upward while failure

spiraled downward. As an explanation to this phenomenon, small businesses’

competitiveness typically depended on innovation resourcefulness while larger

organizations emphasized a resource-based view minimizing the innovation risk

(HewittDundas, 2006). In this context, small businesses that offered technology-based,

knowledge-intensive business services (T-KIBS) or large network-based service achieved

innovation based more on professional knowledge while other small business service

firms conformed to supplier-driven innovations (Miles, 2008). Moreover, small

businesses often engaged T-KIBS firms to deploy business technologies such as

enterprise resource planning (ERP) systems as a source of innovative competitive

advantage (Sledgianowski, Tafti, & Kierstead, 2008). Ultimately, developing new

products and services was among the highest innovation concerns for all firms including

small businesses (Vermeulen, Van Den Bosch, & Volberda, 2007).

Small Business IT Innovation

The literature review included a survey of the current understanding relative to

small business IT innovation focus as well as Internet IT innovation practices. Crespell

and Hansen (2008) identified that small business innovation was often achieved by

adoption or synthesis of existing technology resulting in a new process, capability,

product, or service for the firm. Crespell and Hansen further recognized that the small

businesses’ propensity to innovate was predicated in their organizational culture and work

climate. Organizational culture generally was descriptive relating to the common set of

shared meanings among employees about goals, problems, and practices while work

climate typically was contextualized as the organization’s policies and routines as

perceived by employees (Kaplan & Norton, 2004). Moreover, Benitez-Amado,

LlorensMontes, and Perez-Arostegui (2010) analyzed the relationship among market

performance, technological IT and managerial IT resources, and intrapreneurship culture,

which embodied a working environment infectious with creativity, innovation, and

entrepreneurship. Benitez-Amado et al. found that IT resources had a positive effect on

intrapreneurship that in turn was a predictor of firm market performance indicating

competitive advantage.

Additionally, Qureshil et al. (2009) reported that SMEs adopted IT and ICT to

enable competitive advantage evidencing that IT-dependent initiatives resulted in creating

and sustaining value. Qureshil et al. further suggested that the ICT competitive

advantage was applicable to micro enterprises as well. Although McAfee and

Brynjolfsson (2008) argued that the link between technology and competition has become

much stronger since the mid-1990s, they concluded from their research that (a)

IT sharpened competitive distinction, (b) executive leadership positively correlated with

IT-enabled value, and (c) IT-intensive competitiveness was continuous. As further

evidence of IT-enabled firm value, Oh and Pinsonneault’s (2007) empirical study related

to resource-centered and contingency-based IT strategies indicated that investments in

growth-oriented IT applications exhibited a positive correlation to firm profitability,

which is at the heart of competitive advantage. The literature further revealed specific

uses of IT for operational efficiency and focused strategic initiatives.

Specifically, Archer, Wang, and Kang (2008) researched the barriers related to the

adoption of online supply chain management by Canadian SMEs. Archer et al.

recognized the benefits of using shared IT capabilities by small businesses for effective

supply chain decisions illustrating the use of IT innovation to create upstream and

downstream collaboration for competitive advantage. Correspondingly, Al-Qirim (2007)

examined the adoption of e-commerce, communications, and application technologies

innovation by small businesses in New Zealand. Al-Qirim’s research indicated that the

small business executive's Internet IT innovativeness was a primary determinant for

adoption of external-e-mail, intranet, virtual private network (VPN) extranets, and

websites. Similarly, unprecedented acceleration of competition within the U.S. economy

was triggered by the mid-1990s mainstream adoption of the Internet, enterprise software,

and other IT innovations (McAfee & Brynjolfsson, 2008).

E-business, which comprises e-intelligence, e-commerce, and e-collaboration, has

become an accepted process capability with more than 45% of Canadian firms possessing

some form of e-business capability (Raymond & Bergeron, 2008). For example, the

emergence of Internet banking capabilities has revolutionized the IT innovation strategies

of banks and financial industry firms striving for competitive market advantage in

developed and many developing countries (Al-Hajri & Tatnall, 2008). While ecommerce

has become a basic banking staple, Al-Hajri and Tatnall (2008) indicated that Internet

banking adoption is attributed to perceived ease of use, functionality perceived as a

relative advantage, projected operational efficiency improvements, and anticipated

customer relationship enhancement. Employing a strategic typology that included

prospectors, analyzers, and defenders, Raymond and Bergeron (2008) found positive

performance outcomes for small business manufacturers in terms of growth, productivity

and financial performance when the firm’s e-business capabilities was in alignment with

the firm’s strategy. In essence, the e-business innovation potential resided in strategically

aligned IT capability versus merely in the technology innovation.

Additionally, Dibrell et al. (2008) recognized that small businesses were

sophisticated adopters of IT as a competitive tool to facilitate strategy and build core

competencies. Evidencing this sophistication, Chalhoub (2010) investigated the adoption

of innovative practices undergirded by intellectual capital and social networks developing

a theoretical model based on the relationship between technological and process

innovation as well as leadership performance. Cenfetelli, Benbasat, and Al-Natour

(2008) further suggested shifting the firm’s IT strategy from internal management tools to

customer-directed Internet applications (i.e., business-to-customer e-commerce) to

facilitate innovation of core competencies. Correspondingly, Doherty and Terry (2009)

studied the relationship of IT resources and IT capabilities, which were described in

constructs of tangible and intangible IT assets, yielding sustainable competitive

advantage. Doherty and Terry found that outside-in, customer driven IT initiatives, and

spanning, complementary internal and customer driven IT initiatives, were highly

correlated to sustainable improvements in competitive positioning. Although Dibrell et

al. (2008) concluded that synergistic integration of technology with product and process

improvements enhanced small business performance, limitations in their study suggested

additional research to show the correlation of Internet IT innovation to small businesses’

sustainability.

Cloud-Computing Ecosystem

The literature was investigated to discern the rapidly evolving presence of the

cloud-computing ecosystem as a foundation for exploring its potential use in fostering

small business innovations. The conceptualization for cloud-computing innovation has

been traced back to 1969 with the pioneering contributions of Leonard Kleinrock, a chief

scientist with the Advanced Research Projects Agency Network (ARPANET) that

incubated the Internet (Buyya et al., 2009). The Internet, commonly portrayed as the

cloud, has evolved into a global communication network. The Internet provides the

connectivity infrastructure for shared hardware and software technologies as well as

enables ubiquitous access to these technology assets. The deployment of scalable, rapidly

provisioned, and metered IT assets and computing capabilities ubiquitously accessible via

the Internet is known as cloud computing (Cheng, 2010; Smith, 2009). In other words,

cloud computing was portrayed as the manifestation of “the long-held dream of

computing as a utility” (Armbrust et al., 2010, p. 50) that was proclaimed “the new

frontier of the Internet era” (Etro, 2009, p. 179).

The National Institute of Standards and Technology (NIST) depicted cloud

computing in the context of five essential characteristics, three service components, and

four deployment approaches (Mell & Grance, 2010). The cloud-computing essential

characteristics were featured as on-demand self-service, broad network access, resource

pooling rapid elasticity, and measured services. The service components consisted of

software as a service (SaaS), platform as a service (PaaS), and infrastructure as a service

(IaaS). These service components were typically organized in a layered service oriented

architecture whereby IaaS was the lowest layer comprising hardware and networking

components; SaaS was the highest layer consisting of user interfaces and application

functionally; and PaaS was an in-between layer comprising virtualization, operating

systems, databases, application program interfaces (APIs), and other service platforms.

Last, the deployment approaches included private, community, public, or hybrid cloud

strategies. Each of these cloud-computing aspects was further delineated in the

Definition of Terms section.

Emerging in a seemingly boundless environment, the genealogy of cloud

computing was cited as evolving via the linage of open source software (Dwivedi &

Mustafee, 2010; Sharif, 2010). Open source software was a significant enabling factor in

cloud-computing feasibility attributing to the practicality of its vision as the fifth utility.

Although open source software was a dominant force in IaaS operating systems and PaaS

programming and database environments, open source software is rapidly emerging in

SaaS via open source projects aimed at enabling enterprise IT applications and mobile

technology applications (Dwivedi & Mustafee, 2010; Vescuso, 2010).

Cloud computing is uniquely distinct from super computing capabilities that have

been proprietary to large scale data centers typically funded and controlled by

government, academic research centers, and the largest corporate enterprises. Traditional

computing and IT was designed to support an enterprise scale while cloud computing was

envisaged via a broader Internet scale (Etro, 2009). Cloud computing was based on a

global scale ecosystem model for the ubiquitous rapid provisioning of on-demand,

shared-access, technology resources such as networks, servers, computers, storage,

applications, and other IT services (Katzan, 2008, 2010; Mell & Grance, 2010, 2009;

Truong, 2010).

The cloud-computing ecosystem conceptualization illustrated in Figure 4 was

devised as consumable computing services based on virtual, networked, and recursive

layering of hardware and software technologies (Armbrust et al., 2010; Blaskovich, &

Mintchik, 2011; Etro, 2009; Katzan, 2010). Cloud computing was foreseen as the fifth

utility following water, electricity, gas, and telephony; whereby the Internet functions

analogous to electric or telephone connection and transmission lines (Buyya et al., 2009).

Although Katzan (2008) drew a distinction between cloud computing and traditional

utility computing, the capital investment requirements of large-scale IT capability are

operationalized via the cloud-computing pay-for-use scheme for global rapid

provisioning of seemingly infinite on-demand IT services (Mell & Grance, 2010; Rodero-

Merino et al., 2010).

Figure 4. Cloud-computing ecosystem depicting deployment and service models. The

conceptualization comprised various deployment models (private, community, public,

and hybrid) and service models (SaaS, PaaS, and IaaS). SaaS provides applications and

the user interface; PaaS provides virtualization, operating systems, databases, and layered

interfaces; and IaaS provides servers, networking, data storages, and other computing

hardware. The ecosystem model was adapted from “Effectively and Securely Using the

Cloud-Computing Paradigm,” by Peter Mell and Tim Grance, 2009, National Institute of

Standards and Technology. Published by the United States Federal Government.

While delineating cloud computing as merely hardware-as-a-service (HaaS) and

referencing SaaS and SOA as independent technology innovations, Sahoo (2009)

formulated an IT innovation model for evaluating these emerging technologies based on

three contexts: customer value perspective, industry market sustainability, and technology

intricacies. Counter to Sahoo’s viewpoint, Vouk (2008) asserted that SOA is an essential

underlying and enabling architectural construct of cloud computing functioning in

conjunction with virtualization and workflow management.

Brynjolfsson, Hofmann, and Jordan (2010) suggested that the analogy of

computing services in the clouds (i.e., via the Internet) is not as simplistic as other utility

services. Katzan and Dowling (2010) portrayed cloud computing as a utility

conceptualized by service democratization including ubiquitous availability, optimal

resource sharing, and consumer demand provisioning. Utilities such a water, natural gas,

electricity are typically standardized products delivered to end user consumers catalyzing

other innovations. Cloud computing entails a wide range of services and delivery models

evolving at the rate of Moore’s law; thus, utility-like solutions have not been universally

considered well suited to cloud-computing challenges (Brynjolfsson et al., 2010).

Rodero-Merino et al. (2010) proposed a unique consumer abstraction interface,

named Claudia, for the diverse and evolving cloud-computing ecosystem to address the

utility-like provision challenges. Similarly, Goscinski and Brock (2010) devised the

resources via web services framework (RVWS) to offer a higher-level abstraction of

cloud resource offerings. The RVWS model featured provisioning of cloud computing

based on the current dynamic state and characteristics of cloud services and resources

(Goscinski & Brock, 2010) while Claudia featured mobility amidst the universe of

disparate cloud forms (Rodero-Merino et al., 2010). Goscinski and Brock depicted the

macro level design of RVWS to include dynamic brokering, stateful and dynamic web

service definition language (WSDL) documents, web service resource framework,

dynamic attributes, and resource connectors. Rodero-Merino et al. described Claudia in

the context of a service abstraction layer atop a cloud infrastructure manager.

Beyond Claudia and RVWS, Manzalini, Minerva, and Moiso (2010) envisioned

the CASCADAS (component-ware for autonomic situation-aware communications, and

dynamically adaptable services) project as an autonomic approach for cloud computing.

Relying on complex adaptive systems (CAS) for the self-organizing administration of

network of networks (NoNs), the CASCADAS project was a prototype of the autonomic

communication element (ACE) abstraction that was at the heart of managing the

complexities of the cloud-computing service ecosystem. Manzalini et al. (2010)

described ACE as empowered with the duties of fault finding, configuration,

maintenance, and performance optimization, thus autonomically reducing anomalies and

improving efficiency. These capabilities were considered the DNA for the data revolution

inherent in the next forecasted generation of the Web, which is known as

Semantic Web (Web 3.0).

The feasibility for cloud computing resides in the projected computing capability

and efficiency experienced via economy of scale and scope synergies in response to the

amalgamation of fluctuating demands and commitment levels (Buyya et al., 2009; Katzan

& Dowling, 2010; Truong, 2010). Emphasizing cloud-computing capabilities as a

fundamental change in the management of computing needs, Han (2010) illustrated novel

applications for cloud computing to enhance voluminous capabilities in library asset

management. The economic feasibility of cloud computing was most evident when the

alternative computing strategies require significant initial fixed capital costs, thus

resulting in greater initial economic value (Etro, 2009). The practicality of off-the shelf

rapid deployment of cloud-computing services enhanced its viability and feasibility as

well (Truong, 2010).

The financial advantages of cloud computing have become recognized by chief

financial officers (CFOs) as an enticing business model (Ford, 2010; Marston et al.,

2010). Economies of scale derived from centralized, shared computing resources have

spawned price pressure commoditization deemphasizing business requirement

culminating in undisclosed risks of other service limitations to maintain price

competiveness (Durkee, 2010). However, the use of cloud-computing innovation has not

abdicated business responsibilities centric to intellectual rights, licensing, regulatory, and

privacy compliance.

The service level agreement (SLA) has been designed to disclose risks and hidden

costs, thus leveling the market space based on price performance that is essential for

widespread commercialization of cloud computing (Durkee, 2010). In contrast to smaller

organizations, large organizations with existing IT capabilities were reported as savvier

and having greater negotiating power in garnering business model and end user

experience conformance (Ryan & Loeffler, 2010). However, Hayes (2010) asserted that

dominate cloud-computing service providers are less interested in negotiating the

provisions of their SLA versus merely leveraging its basic market appeal.

Sharif (2010) reported that the commercialization and adoption of

cloudcomputing innovation was not merely based on its economic and shared technology

advantages emphasizing consumer centric drivers. Although cloud computing has

fostered innovations in the IT industry, its innovation potential was anticipated to emerge

through novel business paradigms connecting and engaging people, thus effecting a

virtualized economy (Katzan, 2008; Sultan, 2010). The average transaction in the

virtualized economy typically involved upwards of 20 entities amidst various regulatory

and currency systems imposing increased accountability mandates (Ford, 2010).

Although small businesses traditionally have been late adopters of emerging

technologies, Sharif forecasted that the emergence of the cloud-computing virtual

economy would dissolve IT and other barriers distinguishing large and small businesses.

Various standards such as solution deployment description (SDD), configuration

description, deployment, and lifecycle management (CDDLM), open virtual format

(OVF), and other IT practice specific governance methods have served to enable

cloudcomputing innovations (Rodero-Merino et al., 2010). However, the emerging

cloudcomputing ecosystem subsists as an embryonic evolution pending universal

governance of consumption standards, metering accountability, provisioning capability,

and risk mitigation (Armbrust et al., 2010; Buyya et al., 2009; Vouk, 2008). Although

several large companies have invested in building cloud-computing technology

infrastructures, only a few pioneering firms have entered the market to offer cloud-

computing services (Etro, 2009; Louridas, 2010). Obstacles related to early cloud-

computing adoption have been expressed including perceived proprietary lock-in, data

security risks, software licensing, and fault tolerance concerns (Armbrust et al., 2010;

Louridas, 2010; RoderoMerino et al., 2010).

Russell, Yoon, and Forgionne (2010) attributed cloud-computing availability and

reliability apprehensions to its increased architecture and infrastructure complexity as

well as to the multitude of potential Internet access failure points. Performance modeling

and service level agreements have been used to mitigate these concerns (Russell et al.,

2010). Other issues related to fragmented and dispersed data-based assets have

influenced cloud-computing performance; thus, incubating novel data access

methodologies such as MapReduce facilitating parallel data retrieval, processing, and

storage (Lin & Dyer, 2010; Louridas, 2010).

The perceived benefits and obstacles associated with cloud computing were

extracted from the review of the literature (Armbrust et al., 2010; Bennett, 2009; Braude,

2008; Buyya et al., 2009; Cheng, 2010; Katzan, 2008, 2010; Katzan & Dowling, 2010;

Louridas, 2010; Mell & Grance, 2009; Russell et al., 2010; Schadler, 2009; Truong, 2010;

Vouk, 2008), thus predominately reflecting the perception of industry experts versus the

perception of potential adopters of cloud-computing innovations, in particular potential

small business adopters.

The perceived benefits associated with adoption of cloud-computing innovations

based on the literature review included:

x Cloud-computing services are typically on demand and rapidly deployed. x

Cloud-computing services are provisioned real-time based on

consumerinitiated demand exhibiting elasticity.

x Cloud-computing services offered lower initial capital costs as compared to

alternates.

x Cloud-computing costs are operationalized via pay-for-use metering. x IT

competencies for infrastructure, platforms, and applications are outsourced. x The

Internet provided ubiquitous global access to cloud-computing services.

x Cloud-computing services offered seemingly infinite massive and rapid

scalability.

x Increased economies of scale and scope via pooled shared resource utilization

lower overall cost.

The perceived obstacles associated with adoption of cloud-computing innovations

based on the literature review included:

x Large-scale bandwidth is required for data transfers in data-based

cloudcomputing technology.

x Private or propriety cloud-computing technology may result in data lock-in. x

Reduced availability and dependability concerns due to increased failure points

including Internet reliability.

x Increased vulnerability concerns due to security threats and data confidentiality

breaches in shared ecosystem.

x Heightened noncompliance concerns with existing diverse software licensing

requirements.

Despite the issues related to cloud computing, large organizations are evidencing

early adoption of cloud computing for application such as e-mail, calendaring, human

resource information systems (HRIS), customer resource management (CRM),

ecommerce, and other common utility applications (Bennett, 2009; Schadler, 2009).

Hayes (2010) asserted that the more mission critical an application was considered; the

more likely apprehension was perceived about cloud-computing technology. Moreover,

Hayes reported that a 2009 IDC survey of 691 IT executives across the Asian Pacific

revealed that 11% of the respondents surveyed used cloud-based technology in some

form while 64% of those surveyed perceived significant deficiencies in cloud computing

with the remainder lacking an adequate understanding about cloud computing.

Large corporations such as Google, Amazon, IBM, and Microsoft have

emphasized cloud computing in their innovation strategies as cloud-computing service

providers (Bennett, 2009; Braude, 2008; Buyya et al., 2009; Russolillo & Tibken, 2010;

Smith, 2009). In contrast, large public sector organizations such as the IRS, Veterans

Administration, 2010 U.S. Census, and other government agencies initially have

embraced private cloud computing for rapid deployment of massive scale shared

applications (Mell & Grance, 2009; Sharif 2010; Wyld, 2009). These large government

agencies were portrayed as viewing the advent of cloud computing as an emergency

escape from the evident inadequacies and obsolescence inherent in their antiquated

legacy computer systems and IT applications, which are burdened with massive ongoing

operational costs.

Economic downturns have created business pressures inducing executive

administrative mandates to use cloud-computing technology and have been largely

responsible for federal government agency adoption of cloud-computing technology

(“Federal Govt,” 2010; Kundra, 2010; Mell & Grance, 2009). Only 34% of the federal

agencies surveyed in the “Federal Govt” (2010) study reported lacking familiarity with

cloud computing. Although 14% of those surveyed indicated using some form of cloud

computing, 21% of the cyber-security professionals indicated lacking awareness about

cloud computing (“Federal Govt.,” 2010).

The U.S. Department of Health and Human Services (HHS) decided to implement

cloud-computing innovations in response to administrative mandates to reduce costs and

streamline medical records processing (Chatman, 2010; Kundra, 2010). Health and

Human Services perceived the relative advantage of rapid deployment, scalability, lower

capital costs, and adjunct IT expertise while expressing concern over medical records

security and privacy. Another example, the Virtual Computing Laboratory (VCL) based

at North Carolina State University has established a research-based cloud-computing

pilot project linking its college campuses and other out-of-state IBM virtual computing

initiative member universities (Vouk, 2008). Although the literature revealed significant

advantages for adopting cloud computing including economies of scale, provision

elasticity, and massive-scale computing capability (Davis, 2009; Braude, 2008; Smith,

2009), research related to small business use of cloud computing for innovation is

virtually nonexistent.

Recent studies (Buyya et al., 2009; Etro, 2009; Gens et al., 2010) have predicted

slow initial adoption of cloud-computing innovation followed by aggressive accelerated

adoption and cloud computing utilization. Using a dynamic stochastic general

equilibrium (DSGE) calibrated model augmented with endogenous market structures,

Etro (2009) conjectured, with a measured degree of uncertainty, the forecasted growth in

European GDP because of the macroeconomic impact of cloud-computing innovations.

Etro extrapolated these GDP predictions into the creation of more than 430,000 SMEs

and thousands of new jobs for the 25 European countries studied based on five industry

sectors: manufacturing, wholesale and retail trade, hotels and restaurants, transport

storage and communication, and real estate renting and business activities.

Gens et al. (2010) forecasted similar emergence in cloud-computing

commercialization with global public cloud-computing services projected to reach $55.5

billion by 2014, representing a 27.4% compounded annual growth rate from its 2009 $16

billion level. When compared to the 5% forecast in traditional IT growth, the forecasted

cloud-computing growth rate is more than five times as rapid capturing over 12% of the

total traditional IT market by 2014 and consuming approximately one third of the total

new IT investments, which is anticipated to disrupt extant IT market leadership (Gens et

al., 2010). As cloud-computing adoption becomes more globally widespread, Gens et al.

forecasted a reduction from the U.S. dominance in public cloud-computing revenues

measured at more than 70% in 2009 to approximate 51% by 2014 indicating increased

cloud-computing adoption in Europe and developing countries.

Mobile computing devices ranging from the netbook, personal data assistants

(PDAs), and a myriad of smart phones have rapidly become a portal to the utility of

cloud-computing innovations (Greengard & Kshetri, 2010). In developing economies,

mobile computing and cloud-computing innovations have achieved entrepreneurial

economic contribution bypassing the traditional IT development lifecycle. Greengard

and Kshetri (2010) illustrated the development of iPhone apps using virtualization in

Nairobi, Kenya, where iPhone service is not yet available and local computing resources

are limited. Although the relative advantage of cloud computing for small business was

expected to be significant (Buyya et al., 2009; Etro, 2009; Gens et al., 2010; Greengard &

Kshetri, 2010), specificity relative to the adoption of cloud-computing innovation by U.S.

small businesses, in particular Arizona small businesses, for sustained economic vitality

is lacking.

Transition and Summary

The challenges relative to small business sustainability that threatens free

enterprise economy were acknowledged via a discussion of the research problem. The

importance of the gestation and diffusion of IT innovations for small business economic

vitality were recognized in the literature review. Based on its novel business-computing

paradigm, the emergence of cloud-computing innovation has evidenced the potential for

even greater economic value and societal benefit. A deficiency was surfaced via the

literature review in knowing small businesses’ propensity to adopt cloud-computing

technology. This research project has investigated this deficiency by examining the

relationship between small businesses’ perceived cloud-computing innovation attributes

and their adoption intent. This investigation was optimally suited for a quantitative

crosssectional survey design to address the hypotheses stemming from the research

questions postulated. The following Project section is presented to detail more fully the

research method and design, population and sampling scheme, data collection process,

data analysis techniques, and the study’s reliability and validity. Additionally, the study

results and findings as well as their application to professional practice and implication

for social change are presented Section 3.

Section 2: The Project

The Project section was stipulated to delineate the mechanics of the research study

including how the study was conducted, what activities were performed, and other study

parameters. Preceding the discussion of the project’s internal mechanisms, the research

study purpose is revisited to reinforce the emphasis and rationale for this inquiry.

Additionally, the role of the researcher is described in support of the study’s purpose and

in rendering the resultant findings. The specific project inner-workings included the

subject participants, the research method and design, the study population and sampling

scheme, data collection process, data analysis techniques, and the study’s reliability and

validity.

The process of gaining access to, involvement with, and assurance for study

participants’ ethical safeguards is documented in the Participants section. Originating in

the logical problem statement genesis, the distinctiveness of the designated research

method and design are further clarified in the Research Method and Design section. The

population characteristics and study relevance are disclosed in conjunction with its

derivative sampling scheme in the Population and Sampling section. The study’s

measurement instrument, its application, and the items measured as well as the survey

instrument’s reliability and validity are delineated more fully in the Data Collection

subsections. The quantitative analysis of the study data in context with the research

questions and supporting hypotheses are identified and discussed in the Data Analysis

section. Last, the reliability and validity of the study, including the study instrument, are

assessed in the Reliability and Validity section.

Purpose Statement Revisited

The subject area of interest for this inquiry was the propensity of small businesses,

given their economic importance and volatility, to adopt emerging cloudcomputing

innovation. In this doctoral study, the problem researched was the limited early adoption

of cloud-computing innovation by small businesses for improved economic value. The

purpose for this inquiry was to understand the relationship between the perceived

attributes of cloud-computing innovation and the propensity for adoption by small

businesses. Anchored in a predictive DOI theoretical framework, this research was

devised to examine the intention of Arizona small business leaders to use emerging

cloud-computing technology for obtaining the benefits associated with its novel

businesscomputing paradigm.

This objective was accomplished by employing a quantitative cross-sectional

survey research method soliciting input from IT deployment decision-makers from a

sample of Arizona small businesses. Employing Valier et al.’s (2008) survey instrument,

participant information was solicited based on a convenience sample of small businesses

derived from Arizona Commerce Authority’s e-mail list. Valier et al.’s pretested survey

instrument was administered in an online manner via the Internet collecting

measurements of the participants’ perception about cloud-computing innovation attributes

as well as their propensity for cloud-computing technology adoption. The perceptions of

the innovation attributes were equated to the independent variables: compatibility,

complexity, observability, relative advantage, results demonstrable, trialability, and

voluntariness. The dependent variable, intent to use, was framed in the construct of the

participant’s propensity to adopt cloud-computing technology.

The participant survey response data were analyzed using SPSS’s statistical

functionality. Once the survey data were collected, descriptive statistics procedures were

performed to reveal the composition of the study data. Cronbach’s coefficient alpha and

factor analysis procedures were performed to reaffirm the instrument reliability and

validity. Analysis was performed to ensure a normally distributed dataset as a basis for

testing the statistical relationship between each independent variable and the dependent

variable. Bivariate regression analysis was employed to assess the strength of

correlations between the independent variables and the dependent variable. Multiple

regression analysis was employed to assess the strength of relationship between multiple

independent variables and the dependent variable.

The goal of the inquiry, data collection, and data analysis was to understand better

the prospect of cloud-computing innovation adoption by small businesses, potentially

enhancing their economic vitality. The findings from this investigation, which are

addressed in Section 3, highlight the potential for small businesses to capitalize on the

forecasted economic value of cloud computing as well as other societal contributions.

Transcending the benefits of small business economic value, the anticipated societal

contributions included the potential for guiding policy practices for developing employee

job skills, improving private and public sector collaboration, and improving the

environmental consequences of computing.

Role of the Researcher

The involvement by the researcher in the data collection process originated with

the stipulation of the study population and sample criterion based on the research problem

and associated research questions, thus establishing a portion of the study’s boundedness

(Creswell, 2009). Beyond the population sample provisioning, the researcher’s role

during the data collection process in a quantitative study employing a self-administered

survey instrument typically necessitates less direct participant interaction than with other

quantitative research designs or in a qualitative methods study. One of the predominant

characteristics of qualitative research is that the researcher is a key instrument in the data

collection process (Creswell, 2009). Qualitative research is by nature interpretive; it

typically requires the personal involvement of the researcher as an observer or even as an

observer participant. Whereas the quantitative researcher, engaging minimal participant

contact, establishes empirical measurement instruments and procedures engaged in

collecting study data required to deduce analytical conclusions.

Described in detail in the Methods and Design section, this doctoral research

project was modeled after a quantitative cross-sectional survey design using a participant

self-administered Internet survey capability. The role of the researcher during the data

collection process was limited to the following activities:

1. The researcher performed the setup and configuration of the Internet-based

self-administered survey instrument, complete with positive affirmation of

ethical disclosure and consent acknowledgement.

2. By e-mail invitation facilitated via ACA announcement, the researcher solicited

the participants’ anonymous engagement to complete the survey.

3. The researcher retrieved the anonymous survey responses for subsequent

analysis.

4. The researcher responded directly to participant questions and concerns via e-

mail and telephone.

Possessing more than 20 years of experience, the author of this doctoral study has

gained industry expertise in the field of IT strategy. In this capacity, the researcher has

worked directly with IT technologists, senior executives, and boards of directors for small

businesses, large companies, and the public sector as well. Specifically, the researcher

has facilitated and measured increased organizational capacity, improved operational

efficiency, and lowered operating costs accredited to strategically aligned IT innovations.

More information about the researcher was presented in the Curriculum Vitae appended

to this research study. This experience enhances awareness, knowledge, and sensitivity to

many of the challenges and issues centric to strategic IT innovations as well as IT agility

for business continuity. Although conscious effort was made to ensure objectivity, the

possibility of researcher bias existed potentially affecting the data collection and

interpretive analysis activities. Because of the limited researcher role during the data

collection process, the researcher’s experience related to strategically aligned IT practices

was not perceived to pose a material bias to this study.

Participants

The field of interest for this research consisted of Arizona small businesses, which

included firms with fewer than 500 employees and earning less than $50 million

annually. In 2008, Arizona small businesses accounted for nearly all of the state’s

employers and slightly under half of its private sector employees (SBA, Office of

Advocacy, 2009). The preferred individual from each sampled small business to

participate in the online survey was the IT deployment decision-maker, which may be the

chief technology officer (CTO).

Creeger (2009, 2010) recognized the strategic role of the CTO in evaluating,

understanding business application, and adopting emerging technologies. The CTO

typically engages in formal and informal networks sharing and exchanging technology

developments. The CTO’s role was attributed to technology innovation decision-making,

or at a minimum recommending innovation for adoption. The CTO typically has access

to innovation communication channels, is considered a leadership change agent, and is

astutely cognitive of their corporate and industry sector (social system).

Rogers (2003) cited studies in which an individual, such as a director, reported his

or her perception of innovation attributes as a construct for an organizational unit of

measurement and analysis. The CTO’s role uniquely aligned with the five variables

associated with determining the rate of adoption of an innovation, thus qualifying as a

well eligible candidate for participating in this research survey process. However, many

small businesses cannot support an organizational structure that includes a distinct CTO

position. In a small business, the responsibilities of the CTO may be shared by another

area of responsibility. Therefore, the targeted survey participant was the IT deployment

decision-maker, which may be the CTO, chief information officer (CIO), chief executive

officer (CEO), or technology leader equivalent. The study participants were expected to

possess sufficient knowledge of the emerging cloud-computing technology to calibrate its

perceived innovation attributes as stipulated in this study’s Assumptions section.

The ACA cooperated in this study by inviting study participants. Using their

email contact information, ACA issued an e-mail announcement to small businesses for

soliciting participation in this study. Additional specifications regarding the population

and sampling process are stipulated in the Population and Sampling section.

Invited participants were presented information regarding the study’s scope and

purpose as well as opt-in access to the online survey instrument hosted at Survey

Monkey. Disclosure was made to the participants regarding ethical information in

compliance with the Walden University’s Internal Review Board (IRB) requirements

including participant anonymity surety. The researcher’s contact information was

provided to address any participant questions or concerns related to this study.

Beyond the benefit of helping to gain further knowledge related to the research

topic, participants were offered the incentive to receive a copy of the study’s findings.

Optionally, e-mail information was collected at the conclusion of the online survey

process to request a copy of the research results. The optionally entered e-mail

information was maintained in a manner to ensure survey participant anonymity. At the

outset of the survey, each participant was presented ethics and confidentiality information

and was required to affirm participant consent prior to receiving authorization to engage

in the survey. The ethics and confidentiality disclosure information that was presented

online requiring participant affirmation prior to survey entry is featured in Appendix D.

Research Method and Design

A postpositivist worldview undergirded the premise of beliefs guiding the

methodology and design of this inquiry. Based on a deterministic philosophy (Stacey,

2007), this doctoral study was undertaken to understand the relationship of small business

leaders’ perception of cloud-computing innovation attributes and their intent to adopt this

emerging technology. The research method and design were logically derived from the

study’s applied business problem statement. The specific research methodology and

design strategy stipulated for conducting this DBA study is described more fully in the

following subsections.

Method

Creswell (2009) reported that postpositivism thinking challenged positivism, a

prior traditional viewpoint about the absolute truth of knowledge. The postpositivist

assumptions are steeped in deterministic philosophy that emphasizes examining

relationships between items studied (Stacey, 2007). This vantage has been predominately

used for quantitative research, commonly referred to as the scientific method or the

empirical science. Empirical research has been classified as deductive reasoning,

applying generalities to make inferences about the specifics. Deductive reasoning has

been used as a lens for testing hypotheses derived from research questions (Génova,

2010; Stacey, 2007). Alternatively, qualitative methods have been recognized as more

exploratory and inductive, formulating general rules based on particular cases (Creswell,

2009; Génova, 2010).

This inquiry was conceived in a postpositivist deterministic viewpoint by

examining the relationship of small businesses’ perception of cloud-computing

innovation attributes and their intent to adopt the emerging cloud-computing technology.

As such, a quantitative research method was elected for this research project.

Endeavoring to verify as well as refute hypotheses, quantitative researchers accept only

reasonable hypotheses possessing significant explanatory power (Génova, 2010). The

study’s hypotheses were empirically tested to address specific inferential research

questions derived from the problem statement.

The purpose of the research was unfolded in clarity through the framing

parameters of the research method elected (Creswell, 2009). Using the lens of Valier et

al.’s (2008) theoretical model related to the adoption of IT during the prediffusion stage

of an innovation, this quantitative study investigated the relationship between the

independent study variables and the dependent variable. The quantitative analysis

specified in this study employed SPSS’s statistical functionality to perform descriptive

and regression analysis computations based on the dependent variable, small business

leaders’ intention to use cloud-computing technology, and seven independent variables

representing its perceived innovation attributes: compatibility, complexity, observability,

relative advantage, results demonstrable, trialability, and voluntariness. The quantitative

research method was ideally suited for deductively testing the study’s objective theories

by empirically measuring and statistically analyzing variables to understand construct

relationships (Campbell & Stanley, 1963; Creswell, 2009; Génova, 2010; Onwuegbuzie

et al., 2009; Stacey, 2007).

Offering an alternative approach to quantitative inquiry, qualitative research

methods are typically framed in the embrace of social constructivism, interpretivism,

critical theory, or advocacy and participatory philosophical perspectives while mixed

methods research is adopted based on pragmatism as a worldview (Creswell, 2009;

Onwuegbuzie et al., 2009). Qualitative research procedures are contextualized by

formative and interpretive narrative inquiry, observation, and analysis activities.

Qualitative research methods are generally exploratory in nature addressing open-ended

questions inductively seeking to construe the meaning about a phenomenon derived from

participants’ vantage resulting in findings of particularity versus generalizability

(Creswell, 2009; Yin, 2009). Qualitative case study research has been illustrated as an

effective method for exploring postdictive DOI research to understand the influences

attributed to the nature of social systems or the change agents (Rogers, 2003). Although

distinct case themes are reflective of the particularity of the qualitative generalizations

(Creswell, 2009), the rationale for the multiple case design is derived directly from the

literal and theoretical replications from case to case, thus reporting the interpretive

findings of comparative cases (Yin, 2009). The personal researcher time and effort

intense qualitative exploratory methods typically are not as well suited as quantitative

methods to predictive or deterministic focused inquiry outcomes.

Mixed methods research is recognized as a combination of the interdisciplinary

strengths of deductive and inductive practices attributed to quantitative and qualitative

epistemologies, respectively (Cassell et al., 2006). Mixed methods research has grown in

popularity in social and human sciences studies (Creswell, 2009). The inherit

characteristics of the qualitative aspect of a mixed methods approach were deemed to

necessitate an expansion of the study’s purpose statement and research problem.

Although mixed methods research may have rendered applicability related to this inquiry,

the scope of a mixed methods approach far exceeded the boundedness set forth by the

research problem definition.

Research Design

Quantitative studies are acknowledged as offering fewer design approaches as

compared to qualitative research methods. Creswell (2009) recognized the existence of a

multitude of qualitative research approaches that he has distilled into the five distinct

inquiry strategies of narrative, phenomenology, ethnography, case studies, and grounded

theory. In contrast, quantitative research typically has been classified as either survey or

experimental methods (Creswell, 2009). Campbell and Stanley (1963) featured 16

distinct experimental and quasi-experimental design models while Creswell featured two

nonexperimental survey design strategies: cross-sectional and longitudinal.

While reiterating the importance of the rigor associated with the design of

experiments, Campbell and Stanley (1963) recognized that ultimately every experiment

was imperfect and fraught with potential sources of invalidity. Resembling other rigorous

aspects of experimental design, the quasi-experimental design typically has lacked the

element of random assignment and full control over scheduling of data collection

(Campbell & Stanley, 1963; Trochim, 2006). The nonexperimental survey design was

recognized as an alternative quantitative inquiry approach for examining trends, attitudes,

or opinions via a sample of the subject population (Creswell, 2009). Claims or

generalizations about the population typically are deduced based on the survey sample

results.

Conceived in a survey design in which the study sample is derived from the entire

sampling frame, this research was fashioned from a traditional quantitative crosssectional

survey design strategy that was not intended as an investigation of causation. Creswell

(2009) described the quantitative cross-sectional survey design to include four elements:

literature-based theoretical hypotheses and study variables, population sampling criteria,

a single point data acquisition process via a survey instrument, and statistical data

analysis and interpretation. The distinction between a cross-sectional and a longitudinal

survey design strategy resides in the data collection process; whereby the longitudinal

survey design strategy collects sample measurements at multiple points during an

extended period (Creswell, 2009; Davidsen & Krogstie, 2010).

Bardhan (2007) and Davidsen and Krogstie (2010) illustrated the distinction

between the cross-sectional and longitudinal survey research design strategies. Using a

longitudinal survey research approach, Davidsen and Krogstie studied intrinsic problems

and aspects related to information systems development and support by comparing survey

results collected in 2008 with prior surveys conducted during a prior 15-year period.

Conversely, Bardhan demonstrated the use of a quantitative cross-sectional survey design

strategy. Based on a one-time survey of development product managers, Bardhan’s

research tested the development of adaptive structuation theory (AST) in relation to

collaborative product commerce (CPC) practices associated with IT innovation.

Cloud computing was featured as a newly emerging IT innovation (Buyya et al.,

2009; Etro, 2009; Katzan, 2010; Truong, 2010). The rationale for using a cross-sectional

survey design approach was anchored in providing statistical inferences from

measurement data collected at a specific point versus using multiple survey results

collected during a longer period. Using the cross-sectional survey design achieved the

predictive research problem requirement by measuring the participants’ perceived

innovation attributes in conjunction with their intent to use the innovation prior to or

concurrent with adoption (i.e., prediffusion).

Quantitative survey data typically is collected via a self-administered survey or,

alternatively, by conducting structured interviews with the participants (O'Hegarty et al.,

2010). The structured interview procedure is often used when the researcher is seeking to

collect other observable data in the process or when the self-administered surveys are not

suitable for sample participants use (Creswell, 2009; O'Hegarty et al., 2010). The

selfadministered quantitative survey design approach was preferred over a structured

interview strategy to more efficiently and cost effectively measure the requisite sample

size in a timely manner (Bardhan, 2007; Creswell, 2009). Moreover, this cross-sectional

research design entailed using a web-based self-administered data acquisition process

hosted by the online survey provider Survey Monkey. Additional, information about the

survey instrument, the survey administration, and the survey data retrieval have been

more fully discussed in this study’s Data Collection section.

Web-based survey tools have been used in numerous IT related surveys and offer

benefits over traditional survey practices (Maronick, 2009; Oke et al., 2007; Valier et al.,

2008). The use of a web-based survey tool is preferred due to responses being received

more quickly than with traditional mail methods. Costs were calculated lower using the

online survey tool because printing, mailing, and handling expenses were mitigated. The

use of a web-based survey tool has benefited the participant by ensuring their anonymity

while capturing necessary informed consent, eliminating manual response remittance

activities, and assimilating with modern practices in electronic communication methods.

The population and sampling scope was an essential factor in informing this study

research method and design. Originating from the study’s problem statement, the

theoretical population was described in the context of small businesses. The study

population, which is the accessible population for collecting the study sample, was

further honed by the purpose statement stipulating Arizona small businesses. Practical

limitations in compiling an inclusive list of all Arizona small businesses precluded

conducting simple random sampling of the study population in its entirety.

Narrowing the sample scope by the sampling frame of ACA’s small business e-

mail list provided viable access to derive an Arizona small business sample. A further

narrowing of ACA’s e-mail list via randomization was not elected because it was deemed

potentially to yield an insufficient sample response. Therefore, a convenience sample

based on the entire ACA small business e-mail list (Creswell, 2009) was used for this

study. See the Population and Sampling section for more specific information about the

population and sampling process performed.

Reciprocally, the research method and design strategy were shaped by and molded

the study’s data analysis process. Guided by the study’s problem statement, the research

design strategy became the mechanism used to narrow the viable data analysis processes

suitable for formulating conclusions related to the study’s research questions. The

research design for this study stipulated quantitative analysis using SPSS’s statistical

functionality to perform bivariate regression and multivariate regression computations

based on seven independent variables and one dependent variable. See the Data Analysis

section following for more specific information about the data analysis process

performed.

Population and Sampling

The population and sampling process was an essential aspect of this study’s

design strategy and for maintaining its integrity by ensuring internal and external validity.

The study’s scope and boundedness was framed by the population and sampling process.

The role of the population and sample was discussed as a prelude to indentifying the

study population and describing the specific sampling method employed. Last, the

rationale and appropriateness of the elected sampling as well as the study participant’s

eligibility has been explained.

In quantitative research, statistical inference is a means of drawing conclusions

about a population based on a subset of the population known as a sample (Anderson,

Sweeny, & Williams, 2008; Creswell, 2009). The quality of the statistical inference is a

function of the sampling method applied for collecting study measurements. Probability

sampling is considered more generalizable for making population inferences than

nonprobability sampling. A precise understanding of the population of interest and the

sampling method employed was crucial for the study’s design and the integrity of its

findings.

Population

The population was described in a tiered context with the theoretical population at

the highest level. The theoretical population typically represents the masses to which the

study may be generalized (Trochim, 2006). The theoretical population has its origin in

the study’s overarching problem, which in this case was stipulated via the applied

business problem statement as small businesses. The study population was defined by

narrowing the theoretical population via the lens of the purpose statement to the domain

of accessible subjects (Trochim, 2006). Representing the total subject domain of interest

from which the sample was drawn, the study population consisted of Arizona small

businesses. The sampling frame consists of the source means for gaining access to the

study population to select a study sample (Anderson et al., 2008; Trochim, 2006). The

sampling frame for this study was designated as ACA’s small business e-mail list. A

conceptualization of the population sampling is presented in Figure 5.

Figure 5. Conceptualization of the population sampling refinement process. The

conceptual funnel was depicted illustrating the distillation of the theoretical population of

small businesses into the purpose statement’s study population of Arizona small

businesses that were accessed via the sampling frame of ACA’s small business e-mail list

for soliciting participation in this study’ survey to designate the study sample.

The U.S. Census Bureau (2008) reported that Arizona small businesses, which

constituted this study’s population, numbered at approximately 106,800 firms in 2008.

An industry breakdown of the study population is depicted in Figure 6 to illustrate the

diversity of the Arizona small business in 2008. However, the top four numbering sectors

of the 19 industry sectors, including construction, professional together with scientific

and technical, health care and social assistance, and retail trade, comprised slightly more

than half of the total Arizona small businesses.

Theoretical Population: Small Businesses

Study Population: Arizona Small Businesses

Sampling Frame: ACA

Small Business Email List

Study Sample: Arizona

Small Businesses

ACA Subscribers

Participant:

Arizona Small

Business Leader

Figure 6. Bar chart of Arizona small businesses by industry sector. The number of firms

within each industry sector was extracted from “Statistics of U.S. Businesses (SUSB):

States, NAICS sectors,” published by U.S. Census Bureau in 2008.

The ACA was a newly established quasi-public state authority replacing the

Arizona Department of Commerce governmental agency. The ACA’s purpose has been

described as creating vibrant communities and a globally competitive Arizona economy

via collaboration among private and public sectors to stimulate business startups, foster

small business growth, and attract new businesses to the state (ACA, 2010). The ACA

has maintained an e-mail list that it estimated to contain contact information for

approximately 4,500 small businesses at the time of sampling frame election. ACA’s e-

Arizona Small Businesses by Industry Sector

9,094

5,343

142

10,707

6,703

154

346

324

1,179

12,446

5,211

1,590

1,553

187

7,107

7,776

15,163

15,262

Other Services

W holesale Trade

Utilities

Transportation and W arehousing

Retail Trade

Real estate and Rental & Leasing

Professional, Scientific, & Technical

Mining, Quarrying, Oil & Gas

Manufacturing

Management of Companies

Information

Health Care and Social Assistance

Finance and Insurance

Educational Services

Construction

Arts, Entertainment, and Recreation

Agriculture, Forest, Fishing & Hunting

Administrative, W aste & Remediation

Accommodation and Food Services

mail list represented a cross-section of Arizona firms from the various industry sectors

and was used as the sampling frame for this study.

Sampling

Compiling a list of the entire population of Arizona small businesses was

impracticable. In lieu of soliciting survey participation from the entire study population,

Creswell (2009) recommended probability sampling by collecting survey measurements

from a portion of the population for inference about the entire study population.

Probability sampling consists of the rigor to ensure that the likelihood of being selected is

known for each subject participant (Anderson et al., 2008). An alternative and less

desirable method is nonprobability sampling, which consists of convenience, purposive,

or judgment sampling.

In this instance, the sampling frame consisted of an ACA maintained e-mail list

containing small business leaders’ e-mail addresses. In response to opt-in and opt-out

requests, ACA’s e-mail list fluctuates. At the time of the survey announcement, 3,897

survey participation announcements were distributed by ACA using its e-mail list. A low

response rate of roughly 3% was anticipated based on prior similar research results

(Maronick, 2009; Oke et al., 2007). In lieu of inviting a randomized subset of the sample

population, the entire ACA e-mail list was solicited for participation in the survey,

ensuring a sufficient number of participants to derive meaningful results. Accordingly, a

convenience sample method using the entire sampling frame was elected for this study.

Macht and Robinson (2009) advocated convenience sampling with sufficient sample

responses as opposed to insufficient sample responses from a random sample. Based on

the sampled Arizona small businesses, the individual targeted as the survey participant

was designated as the technology deployment decision-maker, which may be the CEO,

CIO, CTO, or technology leader equivalent.

Rationale, Appropriateness, and Eligibility

Recapping the population and sampling frame, the targeted survey participant was

the individual responsible for IT deployment decision-making within an Arizona small

business. Qualified participants were invited to take part in the online study survey via e-

mail announcements facilitated by ACA based on its small business e-mail listing.

Constituents in the ACA e-mail list, typically, opt-in via e-mail contact (ACA, 2011).

The ACA e-mail list was designated as the sampling frame for this study.

This study was crafted to rely on e-mail as the means for inviting participation in

the online survey. Recent research (Maronick, 2009; Oke et al., 2007; Valier et al., 2008)

demonstrated the viability of e-mail as a vehicle to solicit survey participation. Noted

advantages of e-mail and Internet surveys included low cost, speed, and technology

accepted modality while some disadvantages included low response rate concerns,

questionable quality of expedited responses, and unsolicited e-mail filtering restricting

access (Maronick, 2009). The concern related to the rejection of unsolicited e-mail

filtering was mitigated as the ACA was designated as the sender via its e-mail list service

and the opt-in recipients anticipated e-mails from ACA.

The study sample was initially estimated by ACA to contain of approximately

4,500 Arizona small businesses. This estimated study sample was expected to generate

approximately 135 participants equating to a 3% response rate. The sample size was

estimated due to fluctuations in ACA’s e-mail list of Arizona small businesses. However,

the actual sample size was 3,897, generating 151 completed surveys, which was slightly

lower than a 4% response rate. The survey response rate was well above the required 117

number needed to achieve a medium effect size of 0.30, a significance level of .05, and a

power of .90 as computed by the G*Power version 3.1.2 software program (Faul et al.,

2009). The G*Power program was available free from the Internet via the universal

resource locator (URL) address

http://www.psycho.uniduesseldorf.de/abteilungen/aap/gpower3/.

The study participants were expected to assess the innovation attributes related to

the emerging cloud-computing technology. The CTO’s responsibilities are uniquely

aligned with the five variables associated with determining the rate of adoption of an

innovation: perceived attributes of innovation, type of innovation-decision,

communication channels, nature of social system, and change agent promotional

effectiveness (Rogers, 2003). The CTO’s uniquely aligned responsibilities were

designated as the eligibility qualifications for candidates participating in this research

survey. Not all small businesses could support an organizational structure that included a

distinct CTO position. The responsibilities of the CTO may have been distributed to

another area of responsibility. Therefore, the targeted survey participant was the IT

deployment decision-maker, which may be the CTO or technology leader equivalent.

Ethical Research

Ethical research was a paramount concern for the researcher, Walden University,

study participants, and others who rely upon the outcomes of this research. Ethical

procedures and practices were undertaken ensuring that this research was conducted

ethically. These ethical procedures included a participant consent and withdrawal

process, disclosure of incentives, data safeguard practices, and compliance with Walden

University’s IRB guidelines.

An announcement disclosing the research purpose and soliciting study

participation was issued via ACA using its e-mail publication distribution list. A copy of

the Letter of Cooperation from the ACA was presented in Appendix E. Individuals

electing to participate in the study linked to the survey hosted at Survey Monkey using

the Internet URL provided in the ACA e-mailed invitation announcement.

As a prerequisite to the online survey, participants were issued information

regarding the study’s scope, purpose, participation qualifications, and ethics and

confidentiality disclosure consent requirements. In compliance with the Walden

University’s IRB requirements, disclosure was made to the participant regarding ethical

information including participant anonymity surety. The IRB approval number issued by

Walden University for this research study was 07-07-11-0160136. Researcher contact

information was provided for directly addressing any participant rights or privacy

questions or concerns. Additionally, contact information to a Walden University

representative was provided for other participant ethics concerns.

Participants were given the option to print a copy of the ethics and confidentiality

disclosure. After reading the ethics and confidentiality disclosure, participants electing to

proceed with the survey were required electronically to affirm agreement and consent

prior to receiving access to the online survey instrument. Each participant was advised

that at anytime throughout the survey process he or she may cancel the survey and

withdraw his or her participation from the study. Moreover, if a participant considered

any survey question too personal he or she was advised that he or she may elect to skip

the question. After completing the survey, each participant was prompted optionally to

provide, at his or her sole option, an e-mail address for receiving a copy of the completed

study results.

The ethics and confidentiality disclosure statement was written stipulating that

participant incentives were limited exclusively to the personal gratification of

involvement in the study outcomes and the opportunity to receive, at the participant’s sole

discretion, a copy of the completed study results. No compensation or other inducements

were offered for study participation. A copy of the Ethics and

Confidentiality Disclosure Consent form is included in Appendix D.

Data collected through the online survey were maintained in a confidential and

safe manner. No names of individuals or organizations or other identifying information

were collected or stored with the survey data. Optional e-mail addresses entered

requesting a copy of the study results were stored in a repository separate from the survey

data and were used exclusively for that purpose. The survey data were collected to

perform statistics functions as stipulated in the Data Analysis section. A copy of the

study data was encrypted and maintained in a safe location for a minimum of 5-years to

protect the rights of participants. A signed copy of the Researcher Confidentiality

Agreement form is included in Appendix F.

Data Collection

The tools and the techniques crafted to acquire the prescribed data that was

applied to the ensuing analysis for testing a priori hypotheses was the emphasis of the

research data collection. The data acquisition tool is synonymous with the study survey

instrument (Creswell, 2009). The data acquisition activities included the data collection

and data organization techniques. Additionally, operational definitions for the

quantitative study variables are presented as a prelude to the discussion of the instrument

and data acquisition activities.

Operational Definitions of Variables

The relationship of small business leaders’ perceived attributes of innovation

relative to cloud computing and their intention to use the cloud computing encompasses

the conceptualization of the constructs measured in this study. The construct perceived

attributes of innovation comprised seven independent variables: relative advantage,

compatibility, complexity, observability, trialability, results demonstrable, and

voluntariness. The construct intent to use constituted the dependent variable. The

elements used to assess each of these constructs were derived from prior research (Valier

et al., 2008). The validity and reliability of these elements has been substantiated in prior

peer reviewed research (Moore & Benbasat, 1991; Rogers, 2003; Valier et al., 2008).

These variable elements provided a reliable and effective means for this study’s data

collection and measurement process. Each variable was operationalized by construct

definition and element composition. Each item was measured using a 7-point Likert-type

scale of ordinal values ranging from 1 to 7, with 1 = strongly disagree, 4 = neutral, and 7

= strongly agree.

Compatibility: Independent variable (X1). Anchored in Rogers’s (2003)

compatibility construct assessing the perceived degree of congruence attributed to the

innovation based on the adopter or potential adopter’s experiences, values, and needs,

Valier et al.’s (2008) operationalized definition was applied as the compatibility

measurement for this DBA study. Ensuring construct reliability as well as

appropriateness, Valier et al.’s compatibility measurement, adapting for cloud computing,

comprised Items 1 through 4 as stipulated in the sample PreDOI survey instrument

featured in Appendix A. Compatibility was an independent variable in the context of this

study.

Complexity: Independent variable (X2). Valier et al.’s (2008) complexity

operationalization, which was based on Rogers’s (2003) construct assessing the perceived

degree of difficulty or simplicity attributed to understanding and using the innovation,

was applied as the complexity measurement for this study. Ensuring construct reliability

as well as appropriateness, Valier et al.’s complexity measurement, adapting for cloud

computing, comprised Items 5 through 10 as stipulated in the sample PreDOI survey

instrument featured in Appendix A. Complexity was an independent variable in the

context of this study.

Intent to Use: Dependent variable (Y). Derived from Rogers’s (2003) intent to

use construct assessing the subject’s propensity to adopt an innovation based on its

prediffusion information, Valier et al.’s (2008) operationalized definition was applied as

the intent to use measurement for this DBA study. Ensuring construct reliability as well

as appropriateness, Valier et al.’s intent to use measurement, adapting for cloud

computing, comprised Items 36 through 39 as stipulated in the sample PreDOI survey

instrument featured in Appendix A. Intent to use was a dependent variable in the context

of this study.

Observability: Independent variable (X3). Valier et al.’s (2008) observability

operationalization, which was derived from Rogers’s (2003) construct assessing the

perceived degree of visibility to others attributed to the results of the innovation, was

applied as the observability measurement for this study. Ensuring construct reliability as

well as appropriateness, Valier et al.’s observability measurement, adapting for cloud

computing, comprised Items 11 through 14 as stipulated in the sample PreDOI survey

instrument featured in Appendix A. Observability was an independent variable in the

context of this study.

Relative advantage: Independent variable (X4). Valier et al.’s (2008) relative

advantage operationalization, which was based on Rogers’s (2003) construct assessing

the perceived degree of betterment attributed to an innovation in lieu of precursor ideas,

was applied as the relative advantage measurement for this study. Ensuring construct

reliability as well as appropriateness, Valier et al.’s relative advantage measurement,

adapting for cloud computing, comprised of Items 15 through 22 as stipulated in the

sample PreDOI survey instrument featured in Appendix A. Relative advantage was an

independent variable in the context of this study.

Results demonstrable: Independent variable (X5). Based on Moore and

Benbasat’s (1991) results demonstrable construct assessing the perceived degree to which

tangible evidence of the benefits of an innovation is communicable, Valier et al.’s (2008)

operationalized definition was applied as the results demonstrable measurement for this

study. Ensuring construct reliability as well as appropriateness, Valier et al.’s results

demonstrable measurement, adapting for cloud computing, comprised Items 23 through

26 as stipulated in the sample PreDOI survey instrument featured in Appendix A.

Results demonstrable was an independent variable in the context of this study.

Trialability: Independent variable (X6). Based on Rogers’s (2003) trialability

construct assessing the perceived ability for use inspection attributed to the innovation,

Valier et al.’s (2008) operationalized definition was applied as the trialability

measurement for this study. Ensuring construct reliability as well as appropriateness,

Valier et al.’s trialability measurement, adapting for cloud computing, comprised Items 27

through 31 as stipulated in the sample PreDOI survey instrument featured in Appendix A.

Trialability was an independent variable in the context of this study.

Voluntariness: Independent variable (X7). Valier et al.’s (2008) voluntariness

operationalization, which was based on Moore and Benbasat’s (1991) construct assessing

the perceived degree of self-determined sovereign choice to use an innovation, was

applied as the voluntariness measurement for this study. Ensuring construct reliability as

well as appropriateness, Valier et al.’s voluntariness measurement, adapting for cloud

computing, comprised Items 32 through 35 as stipulated in the sample PreDOI survey

instrument featured in Appendix A. Voluntariness was an independent variable in the

context of this study.

Instrument

The discussion about the instrument was conducted to address its composition and

integrity. The instrument source and the specific data collected were defined via the

instrument composition. The instrument integrity was examined to reveal the

instrument’s reliability and validity.

Instrument composition. With the authors’ permission as evidenced in

Appendix G, the instrument used in the acquisition of the study data was adopted from

Valier et al.’s (2008) PreDOI survey instrument. Change to the instrument was limited to

replacing the reference of open source software to cloud computing throughout the

PreDOI survey instrument leaving the remainder of the survey instrument unaltered.

Regardless, cloud computing was conceived as having an evolutionary lineage of open

source software; conceptually cloud computing has been viewed somewhat synonymous

with open source software (Dwivedi & Mustafee, 2010; Sharif, 2010; Vescuso, 2010). In

many facets of cloud computing such as IaaS and PaaS as well as with recent trends in

SaaS, open source software was considered a primary enabler (Dwivedi & Mustafee,

2010).

Previously used for examining prediffusion adoption of IT innovations, Valier et

al.’s (2008) PreDOI survey instrument was easily applicable to this study’s data collection

requirements. As with Valier et al.’s study, the PreDOI survey instrument employed in

this study was administered via a hosted online web-based platform, which was Survey

Monkey. Study participants were invited by e-mail and directed to an Internet URL link

(www.surveymonkey.com/s/sepowelson) to contribute their survey input. A depiction of

the PreDOI survey instrument highlighting the prompts for data item input was presented

in Appendix A. The PreDOI survey instrument contained two main sections: general

participant items and cloud computing research items.

The general participant information was derived in total from 16 selectable

response items capturing nominal, ordinal, and interval measurements. The general

participant information included three categories: participant demographics, social

systems, and channel communications. The participant demographics information

included Items A through E, which were constructed to capture the participant

characteristics. The social systems information included Items F through K, which were

configured to identify the context in which innovation was perceived. The channel

communications included Items L through P, which facilitated recording the basic means

of learning about the attributes of innovation.

The cloud computing research section of the PreDOI survey instrument was

constructed to contain measurement scales related to the seven perceived attributes of

innovation variables as well as the intent to use variable. These scales reflected the seven

independent variables and the dependent variable, which were operationalized in the

Operational Definition of Variables subsection. The cloud computing research section of

the PreDOI survey instrument facilitated collecting data from 39 items based on a 7-point

Likert-type scale of ordinal values for each item ranging from 1 meaning strongly

disagree to 7 meaning strongly agree.

The constructs measured in this section of the instrument reflected the

participant’s perceptions about the attributes of cloud-computing innovation and their

propensity to adopt cloud-computing technology. Recognizing that reversely worded

items required inverting, the hypotheses were stated to suggest that the greater the score

for the scale items compatibility, observability, trialability, relative advantage, results

demonstrable and voluntariness ,the more favorable the participant perceives the

innovation attribute while the lower the score for the scale item complexity, the more

favorable the participant perceives the innovation attribute.

Instruments integrity. The online survey instrument was made available to

Walden University’s IRB for inspection and approval related to ethical compliance

standards. The IRB ethical critique was essential for ensuring internal consistency via

participant anonymity as well as assuring the potential for replicable external validity.

Valier et al. (2008) used their PreDOI survey instrument to measure participant

perceptions related to open source software innovation attributes. Valier et al. tested the

reliability and validity of the survey instrument and compared its results against similar

primary diffusion studies. Valier et al.’s benchmarked results assessing their PreDOI

instrument variables are presented in Table H1.

Psychometric theory has outlined the basis for quantitative assessment of the

reliability and validity of abstract theoretical constructs or variables used to capture

measurement items via survey instruments (Tiku & Pecht, 2010). Contrasting physical

experimental research and empirical psychometric research, Tiku and Pecht (2010)

featured an eight-step process for evaluating the survey instrument’s internal consistency,

which in psychometric terms means reliability, and its validity referring to the

generalizability resultant from the measures functioning as intended.

100

Modeling empirical psychometric methods, Camisón and Forés (2010) developed

novel measures of knowledge absorptive capacity constructs: potential and realized

absorptive capacities. Camisón and Forés demonstrated the ability to measure knowledge

absorptive capacity, an essential agent for innovation adoptions, via confirmatory factor

analysis statistics related to the psychometric properties of reliability, validity, and

dimensionality scales. The psychometric methods illustrated by Tiku and Pecht’s (2010)

and Camisón and Forés’s (2010) quantitative research approach, which empirically

validated their study measures, served as a template for the strategies crafted to address

threats to this study’s validity and reliability. Nevalainen, Larocque, Oja, and Pörsti

(2010) further suggested extending univariate and multivariate nonparametric procedures

ensuring clustered and hierarchical data for statistical legitimacy.

Based on the survey data collected, the survey instrument integrity was

reexamined for reliability by performing the SPSS statistic functions Cronbach’s

coefficient alpha and factor analysis on item groups (Green & Salkind, 2008). A detailed

review of reliability and validity confirmation procedures employed to substantiate the

instrument has continued integrity was presented in the Data Analysis Techniques section.

Comparing the confirmatory statistical results, which are depicted in Table H2, against

Valier et al.’s (2008) instrument reliability and validity benchmarks featured in Table H1

served as a continued verification of this survey instrument’s reliability and validity.

Data Collection Technique

The survey instrument was implemented as a self-administered online web-based

survey hosted at Survey Monkey. Once implemented, the online survey instrument was

101

visually inspected by an independent small panel to confirm that the instrument questions

and parameters are matched to the original survey instrument questions and data

collection stipulations. As a preface to the data collection activity, the online survey

instrument was quality assurance tested to validate its functionality and accuracy in

collecting the scale item responses. Afterwards, the online survey data repository was

initialized to ensure that the test data were cleared prior to study participant entry.

Arizona small businesses included in the sampling frame were sent an e-mail

soliciting their study participation in the online survey process. Participant invitees were

informed about the study’s scope, purpose, ethical consideration and consent

requirements, and participation qualifications as well as opt-in access to the online survey

instrument hosted at SurveyMonkey.com. In compliance with the Walden University’s

IRB standards, the ethical consensus, including participant anonymity surety, was

affirmed as a prerequisite step in the online survey process.

The researcher’s personal contact information was provided so that study

questions, issues, and concerns may be addressed directly. The survey participants were

presented the option to cancel the survey at any time during the process. Additionally,

survey participants were offered the opportunity to skip any general participant

information question that they preferred not to answer. At the end of the survey,

participants were provided the opportunity to see their responses and edit them before

final submission.

The self-administered online survey process conformed to the following general

activities:

102

1. Affirmation of IRB consent form information.

2. Entry of the general participant survey information section data.

3. Entry of the Likert-type responses to each of the scale items contained in the

survey instrument.

4. Optionally, review participant survey responses prior to final submission.

5. Entry of optional election to participate in the study incentives, which consisted

of receiving a copy of the study results.

Data Organization Techniques

Daily throughout the online data collection process, the survey process was

monitored for response levels and the entered survey data were archived for safeguard.

The self-administered survey process was made available for a 10-day period. The

survey participant, at his or her sole discretion, could have elected to receive the offered

study incentives, which consisted of a complimentary report of the study results. The

participant’s incentive election information, which was an e-mail address, was stored in a

repository separate from the survey data. The incentive election information was used to

instruct individuals about retrieving a complimentary report of the published research

results.

At the conclusion of the self-administered survey data collection period, the final

survey dataset was downloaded from SurveyMonkey.com for archival and analysis. The

final survey dataset was encrypted and copied to a permanent online archive and a

writeonce compact disk. Subsequently, the final survey dataset was transformed into an

Excel workbook before importing into the windows-based SPSS program. Thereafter, the

103

study’s data were stored and processed in an SPSS native file format. Additionally, SPSS

log files were maintained as an audit trail of the data transformations, computed

variables, and other statistical manipulations. The SPSS datasets and process log files

were added to the research data archives for safekeeping and research integrity.

Data Analysis Technique

Architecturally framed by the research strategy, the data analysis procedures and

techniques were designed to test the study hypotheses that were fashioned from the

research questions. Prior to hypotheses testing, the survey data were examined via

descriptive statistics and analyzed for reliability and validity confirmation. The statistical

analysis conducted for this research project was performed using SPSS’s statistical

procedures.

The software program SPSS was developed as a statistical analysis software

package for the social sciences. The SPSS program was designed to automate the

complex formulas and calculations used in various statistics allowing researchers to focus

on the research analysis without having to master the arduous intricacies of the statistical

functions (Green & Salkind, 2008). Prior research has been conducted using the SPSS

program to calculate correlation coefficients to determine the relationships, if any,

between independent and dependent variables (Green & Salkind, 2008; Ness, 2005). The

SPSS Windows-based program PAWS Statistics GradPack© version 17.0.02 was

employed to calculate the statistical results for this study based on the survey data

collected.

104

Once the survey dataset was imported into the windows-based SPSS program as

described in the Data Organization Techniques section, the study’s data were maintained

and stored in a SPSS native format throughout the analysis process. To facilitate study

continuity, the variables and scale item responses were labeled consistently with naming

references originating from this study’s definitions and the PreDOI survey instrument. As

discussed in the Descriptive Statistics Highlights section, incomplete survey instances

were removed from the sample dataset as a preliminary step. Responses to items framed

in a reversely worded manner were inverted into their respective reciprocal scale prior to

statistical analysis ensuring consistency in item and scale interpretation. The scale items

requiring reciprocal scale inversion included survey items 8, 9, 10, 13, 26, 32, 37, 38, and

39.

For each survey instance, a single value was computed for each of the eight

variable constructs by calculating the mean value of the participant’s responses

corresponding to all items in the variable scale. The variable scale items were delineated

in the Operational Definition of Variable sections. Based on the PreDOI theoretical

framework applied to this research project, the computed variable values were subjected

to statistical procedures using SPSS to examine the relationship of each of the seven

independent variables and the one dependent variable. In addition to the

hypothesistesting procedures, this study’s survey data were subjected to descriptive

statistics and reliability and validity analysis. Collectively these analysis techniques were

designed logically and sequentially to address the research questions.

105

Descriptive Statistics Techniques

Based on the 151 completed survey responses, univariate descriptive statistics

were performed to concisely and precisely depict the variable and item measurement data

by summarizing distributions of scores (Green & Salkind, 2008). Frequency distributions

by number and percent were generated for the nominal and interval measures. The

descriptive statistics for continuous scale measurement type data included count,

minimum, maximum, mean, and standard deviation.

The initial aspect of the data analysis process involved performing univariate

descriptive statistic procedures for the general participant information and the study

measures composed of the independent and dependent variables. The descriptive

statistics output related to the general participant information is presented in Appendix I.

Based on the general participant information collected, frequencies and percentages were

calculated and presented in a summary tabular format describing the participant

characteristics, small business attributes relating to the social system conceptualization of

DOI, and cloud-computing awareness relating to the communication channel

conceptualization of DOI.

The independent and dependent study variable descriptive statistics included an

analysis of the survey questions and the study variable scale constructs, which are

presented in Appendix J. The study variables were based on the 39 cloud-computing

research survey questions. For each cloud-computing research question, a Likert-type

7-point scale was used to indicate the participant’s degree of disagreement or agreement.

These indicator questions were subjected to descriptive statistical analysis to determine

106

their central tendency and dispersion. A summary of the participant’s responses to these

questions is presented in Table J1. A scan of the initial descriptive statistics indicated that

no responses to the cloud-computing research survey questions were missing from the

sample of 151 participant surveys.

The cloud-computing research questions were examined for data extremes, which

evidenced outliers that were recorded as responses to the instrument survey questions.

Each cloud-computing research question was subjected to an exploratory analysis to

examine the distribution of responses to the question. The computed study variables were

formed as the basis for the analysis and relational hypotheses testing stipulated in this

study. The study variables were subjected to descriptive statistical analysis to determine

their central tendency, dispersion, and distribution. A summary of the study variables’

central tendency and dispersion based on the completed survey dataset of 151 instances is

presented in Table J2 and a summary of the study variables’ distribution is presented in

Table J3.

Testing for normality, a stem-and-leaf analysis accompanied by graphical boxplot

depiction were applied to each computed variable construct to identify out-of-bound

responses that may influence the analysis findings. The outcome of the initial normality

assessment using the completed survey dataset is presented as a bloxplot diagram

consisting of a summary of separate variables in Figure 7. An iterative process of

normality testing was performed eliminating outliers for each variable, individually and

collectively. The outcome of this process resulted in a normally distributed dataset such

107

that each variable was normally distributed ignoring the other variable and each variable

was normally at every level of the other variables.

Figure 7. Completed survey dataset boxplot diagram comprising a summary of the

separate study variables. For each study variable, a boxplot was depicted showing the

quartile ranges with median item and item outliers, if any, based on the 151 completed

survey responses.

The normally distributed dataset was expected to yield linear relationships

between the independent variables and the dependent variable (Anderson et al., 2008).

Using the normalized dataset as a basis, a summary of the study variables’ central

tendency and dispersion is presented in Table J4 and a summary of the study variables’

distribution is presented in Table J5. A boxplot diagram consisting of a summary of

separate variables based on the normalized dataset is presented in Figure 8.

108

Figure 8. Normalized survey dataset boxplot diagram comprising a summary of the

separate study variables. For each study variable, a boxplot was depicted showing the

quartile ranges with median item for the 86 survey responses after eliminating outliers.

The number of survey response remaining in the normalized dataset was 86. The

G*Power software program function “Post hoc: Compute achieved power” was

performed for linear regression t tests that resulted in a power of .84 based on the sample

size of 86, a medium effect size of 0.30, and a significance level of .05. The computed

power was lower than the a priori calculated power of .90 anticipated in the study’s

design. The lower power result was considered a potential limitation related to the data

analysis and study findings.

Reliability and Validity Analysis Techniques

Statistical procedures were performed to reaffirm the internal reliability of the

survey instrument. The SPSS reliability analysis function was employed to confirm the

109

instrument reliability via split-half estimates and Cronbach’s coefficient alpha item

analysis for each variable construct scale. Extending the split-half reliability procedure,

Cronbach’s coefficient alpha was utilized to assess the consistency of the scores among

the items of a measurement construct by averaging all possible split-half combinations

(Trochim, 2006). The actual item measurements were assessed without the need to

convert the measurement data to z-scores as all item measurements were consistently

defined by a 7-point Likert-type scale of ordinal values. Variable construct scale

reliabilities ranged from .78 to .95 indicating that they exhibited an acceptable level of

reliability. Although Cronbach’s coefficient alpha ranges in value from 0 to 1, typically

an Alpha equal to or greater than .7 is considered an acceptable level even though some

authorities advocate for a stronger standard of at least .8 (Jiménez-Barrionuevo,

GarcíaMorales, & Molina, 2011; Wheeler, Vassar, Worley, & Barnes, 2011).

The SPSS factor analysis procedure was conducted to assess the measurement

scales validity by performing a discrete analysis on each scale. An examination of

eigenvalues was conducted to confirm that the scale items measured the construct tested.

Based on the survey data collected, a summary of the instrument reliability and validity

assessments is presented in Table H2. Consistency of the questions with PreDOI

instrument was considered important to enable reliance on the reliability and validity

significance results of previous studies as a basis for this study.

Hypotheses Testing Analysis Techniques

This research project was conceived in a nonexperimental quantitative design

examining study variable relationships (correlations) that was not intended as an

110

investigation of causation in contrast to more formal experimental methods investigating

cause-and-effect. Eight hypotheses were set forth to address the study’s research

questions. The hypothesis tests were conducted using SPSS statistical analysis

procedures based on the operationalized variables, which were calculated as the mean of

the scale item values corresponding to the variable constructs. Based on the PreDOI

theoretical framework applied to this research project, the Hypotheses H1 through H7

theorized relationships between each perceived cloud-computing attribute construct and

intent to use cloud-computing technology while Hypothesis H8 theorized the distinction

concerning the relationships between the attributes relative advantage and voluntariness

with intent to use cloud-computing technology. The SPSS regression analysis output for

each of the Hypotheses H1 through H8 is presented in Appendix K.

Analysis H1 to H7. For Hypotheses H1 through H7, separately each of the

independent variables was examined assessing the extent of the relationship between

small business leaders’ perceived cloud-computing innovation attributes and their

propensity for adoption. Bivariate regression tests were conducted to signal the degree of

linear relations between each independent variable of cloud-computing attributes and the

dependent variable intent to use based on the study sample. The bivariate linear

regression equation related the predicted criterion (Ǔ) to each independent variable (X)

via a slope weight (Bslope) for X and an additive constant (Bconstant): Ǔ = Bslope X + Bconstant

(Green & Salkind, 2008).

Whereas this research design was a nonexperimental quantitative cross-sectional

study, the assumptions underlying the significance test were based on the random effects

111

model. Green and Salkind (2008) ascribed the following assumptions underlying the

significance test for the bivariate linear regressions:

1. The predictor and criterion variables are bivariately normally distributed in the

population.

2. The scores on variables are independent of other scores on the same variables

and are derived from a random sample.

The SPSS linear regression procedure was used to test significance by means of

four correlation index variations: the Pearson correlation coefficient (r), the overall

correlation index (R), a squared correlation (R2), and an adjusted squared correlation

(R2adj). A correlation index equal to zero indicated no correlation between the predictor

and the criterion variables. An index with a value below zero indicated an inverse or

negative correlation while a value above zero indicated a positive predictive correlation.

The closer the correlation index value approached either -1 or +1 the stronger the

independent variable predicted the criterion variable (Anderson et al., 2008). The SPSS

output for the bivariate linear regression included correlation coefficients and overall

summary correlation statistics, which are presented in Table K1 and Table K2,

respectively. Additionally, bivariate scatterplots were developed for each test depicting

the occurrence of outliers as well as the existence of possible linear relationships.

Analysis H8. The eighth research question was an inquiry regarding which

perceived attribute of innovation relative advantage or voluntariness has a stronger

relationship to small business leaders’ intent to use cloud-computing technology.

Hypothesis H8 was specified to examine, during the prediffusion stage, whether small

112

business leaders’ perceived the relative advantage of cloud computing is more strongly

related to their propensity to adopt the innovation than their perceived voluntariness of

cloud computing. The multiple linear regression procedure was performed as a test for

Hypothesis H8 to assess the strength-of-relationship index indicating the degree of

correlation between the multiple predictor variables (i.e., the independent variables

relative advantage and voluntariness) and the criterion variable, which in this case was

the dependent variable intent to use. The multiple linear regression equation related the

predicted criterion (Ǔ) to the linear combination of the independent variables and an

additive constant. With two predictors (relative advantage and voluntariness) as

stipulated in hypothesis H8, the regression equation is Ǔ = B1slope X1 + B2slope X2 + Bconstant

(Green & Salkind, 2008).

The assumptions underlying the significance test based on the random effects

model were applied. Green and Salkind (2008) ascribed the following assumptions

underlying the significance test for the multiple correlation coefficients:

1. The variables are multivariately normally distributed in the population.

2. The cases represent a random sample from the population, and the scores on

variables are independent of other scores on the same variables.

The SPSS multiple regression procedure was used to test significance by means of

four correlation index variations: the Pearson correlation coefficient, the multiple

correlation, a squared multiple correlation, and an adjusted squared multiple correlation

(Green & Salkind, 2008). A multiple correlation index equal to zero indicated no

correlations between the predictor and the criterion variables. A correlation coefficient

113

with a value below zero indicated an inverse or negative correlation while a value above

zero indicated a positive predictive correlation. The closer the correlation coefficient

value approached either -1 or +1 the stronger the independent variable predicted the

criterion variable (Anderson et al., 2008). The multiple regression analysis included part

and partial correlations for each predictor and criterion, partialling out the effects of the

other predictors in the regression equation. The SPSS output from the multiple linear

regression procedures included coefficients and t test, regression correlations, and overall

correlation indices with change statistics, which are presented in Table K3, Table K4, and

Table K5, respectively.

Reliability and Validity

Reliability and validity considerations were integral to this study and were

interwoven throughout, commencing with the initial study design and concluding with

study findings and future implications. The reliability and validity issues were centric to

the survey instrument as well as the overall study integrity. Although the meaning behind

these terms varies depending on the research method, the quantitative definitions were

applied for this study.

The survey instrument’s reliability and validity were discussed in the Instrument

Integrity subsection of the Data Collection section and in the Reliability and Validity

Confirmation subsection of the Data Analysis Technique section. The following was

provided as a recap of the survey instrument reliability and validity. The internal and

external validity of the study was examined as well.

114

Reliability

In quantitative research, reliability has been defined as the assurance of the

internal consistency in the instrument item measures as well as in the test administration

and scoring (Creswell, 2009). The survey instrument used in this study was adopted from

Valier et al.’s (2008) pretested PreDOI survey instrument. The basis for the reliability

and validity of Valier et al.’s PreDOI instrument was demonstrated via statistic techniques

including the factor analysis results presented in Table H1. The survey instrument was

administered via an online web-based service.

Once setup, the survey instrument underwent visual inspections, including an

independent small panel review, and quality assurance tests to validate that scores were

accurately recorded. The nature of the online web-based survey data collection procedure

was constructed to ensure reliability by mitigating the variability associated within in-

personal survey administration. After collecting the study data, statistic techniques were

performed as an initial phase of the data analysis procedures to ensure the instrument has

continued reliability. As previously discussed in the Data Analysis Technique section, the

reliability statistical tests included Cronbach’s coefficient alpha item analysis assessing

internal consistency and factor analysis explaining the statistical variation and covariation

among items.

Additionally, the research question hypotheses testing were conducted using

SPSS’s regression techniques as detailed in the Hypotheses Testing Analysis Techniques

section. Testing Hypotheses H1 through H7, the bivariate linear regression procedure

was applied as a means for examining the study variable relationships stipulated via the

115

theoretical framework. Similarly, the multiple linear regression procedure was applied as

the statistical function to test Hypothesis H8.

Validity

Validity in quantitative research has been defined as the assurance of concluding

meaningful inferences from the instrument scores (Creswell, 2009). Threats to

quantitative research validity have included internal issues, external factors, statistical

conclusions, and construct adequacy. Participants with predisposition threaten internal

validity. Internal validity consisted of assuring the veracity of the relationships between

the independent variables and dependent variable while external validity consisted of

ensuring the generalizability of the study variable relationships beyond the contexts of the

study setting (Creswell, 2009; Onwuegbuzie et al., 2009).

The construct validity was ascertained by the adequacy of study variable

definitions and measures whereas the statistical conclusion validity was confirmed by the

accuracy of inferences drawn from the data based on statistical power and underlying

assumptions (Creswell, 2009). Many of the threats associated with the experimental

research design such as history, maturation, mortality, diffusion of treatment, interactions

or selection, setting, or treatment, and other validity threats (Campbell & Stanley, 1963;

Creswell, 2009) were not applicable to the nonexperimental cross-sectional survey design

approach based on extant empirical psychometric research.

Regression and selection posed the greatest potential internal validity threats to

this research. Creswell (2009) described regression validity issues as extreme participant

scores and selection validity concerns as participants with predisposition to certain

116

outcomes. Descriptive statistics and iterative or follow-up statistical functions as

described in the Data Analysis Technique section were conducted to isolate or otherwise

mitigate regression validity issues. The sampling approach was chosen as an optimal

means for increasing the sample size for addressing potential selection validity concern

based on the study population. Additionally, the SPSS factor analysis function, as

previously discussed in the Data Analysis Technique section, was performed using the

collected survey data to reaffirm the construct validity of the pretested PreDOI survey

instrument.

The nature of the cross-sectional survey study mitigated several traditional

experimental internal validity concerns (e.g., history, maturation, mortality, treatment

diffusion, and control group demoralization or rivalry) because of the limited duration of

participant involvement (Campbell & Stanley, 1963; Creswell, 2009). Independent

expert review and critique of the statistical analysis techniques and procedures were

engaged to elevate the study’s statistical conclusion validity. Correct inferences drawn

from the study data to other settings in past or future situations substantiate quantitative

research external validity (Creswell, 2009). This study’s reliability and validity results

were evaluated against Valier et al.’s (2008) results as well as other prior DOI research

assessments.

Transition and Summary

The mechanics of the research study were revealed in the Project section by

detailing the project’s inner-workings and how the study was conducted including its

underlying assumptions. The purpose for this quantitative survey study was to examine

117

the relationship of small business leaders’ perception of its cloud-computing innovation

attributes and their propensity to adopt this emerging technology. The researcher’s role

was focused on ensuring the reliability and validity of the study while rendering the

resultant findings.

The subject participants were defined as Arizona small business IT deployment

decision-makers, which likely were CTOs or equivalent small business leaders. The

sampling frame was designated as ACA’s small business e-mail list. A convenience

sample was drawn from the entire sampling frame to enhance the study sample size as a

tradeoff to a random probability sample. The rationale for the quantitative research

method was explained in context of the research problem and questions. Additionally, the

suitability of the cross-sectional survey design using a self-administered, online webbased

data collection tool was affirmed.

Valier et al.’s (2008) PreDOI survey instrument served as the research instrument,

which was fully described featuring operational definition for the seven independent and

one dependent measurement constructs. The procedures utilized to safeguard the data

and preserve its integrity including encrypted permanent archives were disclosed. The

reliability and validity of the instrument and overall study was affirmed via statistical

techniques, research design practices, and other quality assurance methods.

The data was analyzed using SPSS statistical techniques to ensure comprehension

of the data, reaffirm reliability and validity of the survey instruments, and to test the

research question hypotheses. Bivariate linear regression tests were applied to test each

of the Hypotheses H1 through H7. Similarly, multiple regression statistics were applied

118

to test Hypothesis H8. The data analysis techniques were explained as a basis for

supporting or disconfirming the stipulated hypotheses.

The interpretation of the data analysis and the study findings, including the

research question conclusions, are presented and related to application for professional

practice in Section 3. The study findings and conclusions are bound by the evidence

collected in context of the theoretical framework. Grounded in the significance of the

study foundation and project outcomes, implications for social change are expressed in

terms of tangible improvements for workforce preparedness, environmental

consciousness, and social entrepreneur enablement. A call to action is derived from the

application of the study conclusions to professional practice and the recommendations for

further study. Section 3 is closed by disclosing reflective insights related to the research

project and the summary conclusions and by encouraging future improvement and

benefit.

Section 3: Application to Professional Practice and Implications for Change

The purpose of this quantitative survey inquiry was to examine the relationship

between small business leaders’ perception of cloud-computing innovation attributes and

their propensity to adopt the emerging cloud-computing technology. The previous

sections of this study were composed to establish the foundation and project workings for

a doctoral study with practical applications. Supported by a review of the professional

and academic literature, the Section 1 foundation included a statement of the problem

researched, a description of the research purpose, a summary of the nature of the study,

119

the research questions and associated hypotheses, and the theoretical framework basis as

well as an emphasis about the study’s significance. Section 2 included a discussion of the

project workings that incorporated an understanding about the role of the researcher, the

participants, the mechanics of research method and design, the population and sampling

approach, data collection technique and data organization practice, and the data analysis

techniques as well as an assessment of the study’s reliability and validity.

Building on the research foundation and project mechanics, Section 3 contains the

research results and its significance. Following an overview of the research study, the

researcher presents the study findings in context of the theoretical framework and

research questions, discusses how the research can be applied to professional practice and

its underpinning implications for social change, and provides recommendations for action

and further study. In closing, the researcher shares reflections about the study and

presents a recap of the study and the research conclusions.

Overview of Study

Although the general problem targeted by this study was the slow embrace of

technology innovations by small businesses, the specific problem researched was honed

to small business early adoption of cloud-computing innovation for improved economic

value. The research project’s purpose was founded on a review of the professional and

academic literature emphasizing five topical areas: small business trends and economic

contributions, small business innovation policy, the strategic role of small business

innovation, small business IT and Internet innovation practices, and the emerging

cloudcomputing ecosystems. Based on this foundation, the purpose for this inquiry was

120

to examine the relationship between small business leaders’ perceptions of cloud-

computing attributes and their propensity to adopt the emerging cloud-computing

technology.

The theoretical framework for this study was adopted from Valier et al.’s (2008)

predictive DOI theory, which predominantly was grounded in Rogers (2003) and Moore

and Benbasat’s (1991) DOI research. The predictive DOI framework was embraced to

assess small business leaders’ intent to use emerging cloud-computing technology and

their perceptions about cloud-computing innovation attributes relying on the variable

constructs relative advantage, compatibility, complexity, observability, trialability, results

demonstrable, and voluntariness. Following this theoretical framework, this research

project was constructed based on a quantitative research method using a nonexperimental

cross-sectional survey design strategy.

In response to the problem statement and the study purpose, given small business

volatility and economic importance, the overarching research question guiding this study

was, “What is the propensity of small businesses to embrace the cloud computing novel

business-computing paradigm?” Anchored in the predictive DOI theoretical framework,

the seven perceived attributes of innovation were designated as independent (predictor)

variables and intent to use was designated as a dependent (criterion) variable. For each of

the seven perceived attributes of cloud-computing innovation, a research question and

corresponding hypothesis was stipulated to examine individually the extent of the

relationship between small business leaders’ perceived attribute of innovation and their

intent to use cloud computing. The eighth research question and corresponding

121

hypothesis was stipulated to examine the strength of relationship between relative

advantage and intent to use in comparison to the strength of relationship between

voluntariness and intent to use.

Valier et al.’s (2008) previously tested online survey instrument, which was

designed to assess the adoption of IT during the prediffusion stage of an innovation, was

applied to collect the data for this study. The study’s survey instrument included 16

questions related to the participant and 39 questions related to small business leaders’

perceived cloud-computing innovation attributes and their intent to use cloud-computing

technology. The survey instrument was implemented as a self-administered online survey

hosted at Survey Monkey.

The research survey was announced via an Arizona Commerce Authority e-mail

distribution soliciting participation from senior IT deployment decision-makers based on

a convenience sample of small businesses drawn from an entire e-mail list maintained by

ACA. The survey participant’s involvement included affirmation of IRB consent, entry

of general participant information, entry of the Likert-type responses to each scale item

question, and optional designation of the election to receive a copy of the study results.

At the conclusion of the data collection timeframe, the study’s data were retrieved,

encrypted and archived, and imported into SPSS for statistical analysis to test the

stipulated hypotheses.

The data analysis procedures and techniques performed were architecturally

framed by the study’s quantitative research method and cross-sectional survey design

approach. The software program SPSS was used to calculate the statistical results based

122

on the survey data collected. The current study’s analytical scope included descriptive

statistics, instrument reliability and validity confirmation, and hypotheses-testing

correlation regressions.

The 3,897 e-mails distributed by the ACA announcing the research survey resulted

in 151 completed surveys equaling a 3.87% response rate. Univariate descriptive

statistics were performed using the survey data to depict the central tendency, dispersion,

and distribution of the participant information, item measurements, and variable data.

Based on the participant information collected via the survey instrument, frequencies and

percentages were calculated and presented in a summary tabular format describing the

participant characteristics, small business attributes, and cloud-computing awareness.

Each cloud-computing research question, which used a Likert-type 7-point scale

to indicate the participant’s degree of disagreement or agreement, was subjected to

descriptive statistical analysis to determine its central tendency and dispersion. Similarly,

the study variables, which were calculated based on the mean of the corresponding item

scale values for each instance, were subjected to descriptive statistical analysis to

determine their central tendency, dispersion, and distribution. For each of these analysis

procedures, summary statistics were reported in tabular format while key statistics are

highlighted in the Descriptive Statistics Findings section.

Additionally, statistical procedures were performed to reaffirm the internal

reliability of the survey instrument. The SPSS reliability analysis function was employed

to confirm the instrument reliability using Cronbach’s coefficient alpha item analysis for

each variable construct scale. The SPSS factor analysis procedure was conducted to

123

assess the measurement scales validity by performing a discrete analysis on each scale.

An examination of eigenvalues was conducted to confirm that the scale items measured

the construct tested.

Analysis was performed to test the research hypotheses by using SPSS statistical

procedures to examine the relationship of each of the seven independent variables and the

one dependent variable. As a basis for the hypotheses testing, normality test procedures

were performed to ensure that the hypotheses-testing dataset was normally distributed.

For each hypothesis, statistical procedures were performed assessing the extent of the

relationship, individually, between each perceived cloud-computing innovation attribute

and intent to use cloud computing. The bivariate regression statistics procedures were

applied separately to each of the independent variables and the dependent variable testing

Hypotheses H1 through H7. The bivariate SPSS regression analysis procedures were

performed to assess the degree of relationship, if any, based on each predictor variable

and the criterion intent to use. The multiple regression statistics procedures were applied

to test Hypothesis H8, which assessed the strength of relationship between relative

advantage and intent to use in comparison to the strength of relationship between

voluntariness and intent to use.

The basis for the study findings was framed in the interpretation of the results of

the data analysis procedures. With the exception of Hypothesis H7, the null hypothesis

associated with each research question was rejected; therefore, the findings were

interpreted as supporting the hypothesized predictions for each research question other

than the seventh research question. Statistical diagrams are presented that further

124

supported findings from this study. A further discussion of the data analysis

interpretation and study findings is presented in the Presentation of the Findings section.

Presentation of the Findings

This study’s findings were organized by presenting conclusions addressing each

research question. These conclusions were derived from the interpretation of the

outcomes of the data analysis performed based on the survey data collected as described

in Section 2. To address the research questions, hypothesis-testing regression statistics

procedures were applied separately to the stipulated predictor variables and the criterion

variable intent to use. The variable construct intent to use was operationally defined as

the subject’s propensity to adopt cloud-computing technology. An operational definition

for each predictor variable is restated in the discussion of the corresponding research

question findings.

For Hypotheses H1 through H7, bivariate regression analysis procedures were

performed as a test to determine the degree of relationship between each predictor

variable and the criterion intent to use. For Hypothesis H8, the multiple linear regression

procedure was performed as a test to assess the strength-of-relationship index indicating

the degree of correlation between the multiple predictor variables relative advantage and

voluntariness and the criterion variable intent to use. Reinforcing the conclusions, the

study findings were related to the theoretical framework and the extant literature

regarding effective business practices.

Derived from the completed survey dataset, a scatterplot overlay mapping the

variable coordinate relationships between each cloud-computing attribute and intent to

125

use is presented in Figure 9. Although this overlay diagram does not depict any strong

relationships, the scatterplot overlay was provided as a basis for graphically comparing

the predictor and criterion variable relationships based on the completed survey dataset

with the individual scatterplot diagrams based on the normalized dataset presented for

each research question findings.

Figure 9. Scatterplot overlay mapping the predictor and criterion coordinate relationships.

For each sample response based on the completed dataset of 151 surveys, best-fit lines

were formed to depict the linear relationships between each predictor variable and the

criterion variable intent to use including R2 coefficients for each correlation.

Descriptive statistics highlights were presented as a preface to the

hypothesistesting findings to instill greater understanding about the study participants and

126

the research data collected. Univariate descriptive statistics procedures were applied to

the general participant information collected via this study’s survey. The resultant key

descriptive statistics included participant characteristics, small business attributes, and

cloud-computing awareness. Additionally, descriptive statistics related to the study

variables based on the completed survey and normalized datasets were highlighted.

Descriptive Statistics Highlights

As the initial phase of the data collection process, ACA distributed 3,897 e-mails

announcing this study to solicit survey participation. The e-mail announcement

precipitated 154 survey participants. Three of the survey participants elected to cancel

the survey process resulting in 151 completed surveys equaling a 3.87% response rate,

which exceeded the initial response projection calculations. Based on the completed

survey responses, descriptive statistics highlights are presented about the participant

characteristics, small business attributes, and cloud-computing awareness. Descriptive

statistics highlights are presented about the study variables based on the normalized

dataset that resulted from a refinement of the completed survey responses after

eliminating variable outliers.

Participant characteristics. The descriptive statistics composing the participant

characteristics, which were framed on Survey Questions A through E, are presented in

Table I1. Eighty respondents, equaling 53% of the sample, indicated that the

responsibility classification of information technology best fit them, while 37

respondents, representing 24.5% of the sample, designated the executive classification.

The ages 41-50 years and 31-40 years nearly tied for the top age-range category rankings

127

with 55 participants equaling 36.7%, and 49 participants equaling 32.7% of the

respondents, respectively. Although 11 participants constituting 7.3% of the sample

elected not to specify a gender designation, 106 participants equaling 75.7% of the

respondents designated male as their gender. Ninety-eight participants representing

65.3% of the respondents designated the education level classification of a Bachelor’s

degree. The 11-20 years category was designated most for the number of years of

fulltime experience with 73 responses equaling 48.3% of the sample.

Small business attributes. The descriptive statistics composing the business

attributes, which were based on Survey Questions F through K, are presented in Table I2.

Seventy-one participants equaling 47.3% of the respondents designated the classification

corporation as their firm’s legal structure while 62 participants representing 41.6% of the

respondents designated the category 11-20 years as the range of years that their firm had

been in business. Although the number of industry categories in which the participants’

firms operated was broad, the top three designated categories garnered more than half of

the total responses. Totaling more than half of the industry classifications, professional,

science, and technical, finance and insurance, and education were designated by 38

participants equaling 25.5%, 25 participants equaling 16.8%, and 22 participants equaling

14.8% of the respondents, respectively. As a measure of firm size based on the sample,

82 participants equaling 54.3% designated 10-49 employees as the range for the firm’s

number of employees while 87 respondents representing 57.6% selected the classification

of less than $2 million for the firm’s annual revenue category. Although 124 participants

representing 82.7% of the respondents indicated that their firm did not use

128

cloudcomputing technology, 26 participants designated that their firm used at least one

cloudcomputing technology. Eighteen of the 26 respondents indicated that their firm

used cloud-computing designated hosted application such as e-mail, office, or other

utility functions as the nature of their firm’s cloud computing usage.

Cloud computing awareness. The descriptive statistics composing the cloud

computing awareness, which were derived from Survey Questions L through P, are

presented in Table I3. Eighty-three participants equaling 55.7% of the respondents

designated 1-3 years as the range of years in which the participant had known about cloud

computing. Although four individuals elected not to respond to Question M, 112

participants equaling 76.2% of the respondents designated not attending any cloud

computing presentation. Conversely, 124 participants equaling 82.1% of the sample

indicated having read advertisements about cloud computing. One hundred nine

participants representing 72.7% of the respondents indicated that they had talked with a

colleague about cloud computing. Last, 117 participants equaling 79.1% of the

respondents indicated not having previously used cloud computing.

Study variable measures. The refinement of the completed survey responses

based on the normalized study variables restricted the dataset to 86 survey instances.

This normalized dataset was the basis for the study variable descriptive statistics and the

hypothesis-testing regression statistics. Based on the mean study variables,

generalizations about the sample of perceived attributes of cloud-computing technology

and intent to use cloud-computing technology were developed. On average, participants

indicated that they somewhat less than slightly disagreed that cloud computing was

129

difficult to understanding and use while indicating that they somewhat more than slightly

disagreed that its results were visible to others. Participants, on average, indicated

slightly more than neither disagreement nor agreement that they possessed the ability for

use inspection of cloud computing. On average, participants indicated that they nearly

agreed (i.e., more than slightly agreed but less than agreed) that cloud computing was

perceived a betterment in lieu of precursor ideas and cloud computing was congruent

with their experiences, values, and needs. On average, participants indicated that they

more than agreed that the tangible evidence of its benefits were communicable, using

cloud computing was a self-determined sovereign choice, and they intended to use

cloudcomputing technology.

Research Question 1 Findings

The first research question asked: To what extent does relative advantage, a

perceived attribute of innovation, relate to small business leaders’ intent to use

cloudcomputing services? The variable construct relative advantage was operationally

defined as the perceived degree of betterment attributed to cloud-computing technology

in lieu of precursor ideas. Based on a review of the literature related to the cloud-

computing ecosystem and the theoretical framework, the first hypothesis was framed

theorizing that, during the prediffusion stage, the higher the level small business leaders

perceived the relative advantage of cloud computing the greater their propensity to adopt

the innovation. The null hypothesis H10 was stipulated: No correlation exists between

relative advantage and intent to use cloud computing.

130

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived relative

advantage of cloud-computing innovation. The scatterplot diagram depicting relative

advantage and intent to use, as shown in Figure 10, was presented revealing that the two

variables were linearly related such that as the attribute relative advantage increased the

intent to use increased. The regression equation for predicting the intent to use, given

relative advantage, was

Predicted Intent to Use = .49 Relative Advantage + 3.25. (1)

An intent to use value of 3.25, which reflected the regression equation y-intercept value

when relative advantage theoretically equaled zero, was interpreted as a propensity to use

cloud-computing technology between slightly disagree and neither disagree or agree.

The 95% confidence interval for the slope, .26 to .71, did not contain the value zero, and

therefore relative advantage was significantly related to intent to use.

131

Figure 10. Scatterplot mapping relative advantage and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = 4.28, p < .01. The magnitude of the standardized

correlation coefficient was between medium and large, .42, evidencing that relative

advantage was more than moderately related to intent to use. Approximately 18% of the

variance of the intent to use was associated with relative advantage. As hypothesized, the

H10 null hypothesis was rejected (p < .01), and relative advantage was deemed positively

correlated to intent to use. Therefore, based on these findings the first hypothesis, H1,

was supported. The conclusion related to the first research question was interpreted

based on the statistical analysis findings indicating that relative advantage was correlated

to intent to use in a greater than moderate positive degree.

Research Question 2 Findings

The second research question asked: To what extent does compatibility, a

perceived attribute of innovation, relate to small business leaders’ intent to use

cloudcomputing services? The variable construct compatibility was operationally defined

as the perceived degree of congruence attributed to cloud-computing technology based on

the subject’s experiences, values, and needs. Guided by a review of the literature related

to the cloud-computing ecosystem and the theoretical framework, the second hypothesis

was framed theorizing that, during the prediffusion stage, the higher the level small

business leaders perceived the compatibility of cloud computing with extant factors the

132

greater their propensity to adopt the innovation. The null hypothesis H20 was stipulated:

No correlation exists between compatibility and intent to use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived compatibility

of cloud-computing innovation. The scatterplot diagram depicting compatibility and

intent to use, as shown in Figure 11, was presented revealing that the two variables were

linearly related such that as the attribute compatibility increased the intent to use

increased. The regression equation for predicting the intent to use, given compatibility,

was

Predicted Intent to Use = .28 Compatibility + 4.63. (2)

An intent to use value of 4.63, which reflected the regression equation y-intercept value

when compatibility theoretically equaled zero, was interpreted as a propensity to use

cloud-computing technology between neither disagree or agree and agree slightly. The

95% confidence interval for the slope, .14 to .42, did not contain the value zero, and

therefore compatibility was significantly related to intent to use.

133

Figure 11. Scatterplot mapping compatibility and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = 4.00, p < .01. The magnitude of the standardized

correlation coefficient was between medium and large, .40, evidencing that compatibility

was more than moderately related to intent to use. Approximately 16% of the variance of

the intent to use was associated with compatibility. As hypothesized, the H20 null

hypothesis was rejected (p < .01), and compatibility was deemed positively correlated to

intent to use. Therefore, based on these findings the second hypothesis, H2, was

supported. The conclusion related to the second research question was interpreted based

on the statistical analysis findings indicating that compatibility was correlated to intent to

use in a greater than moderate positive degree.

134

Research Question 3 Findings

The third research question asked: To what extent does complexity, a perceived

attribute of innovation, relate to small business leaders’ intent to use cloud-computing

services? The variable construct complexity was operationally defined as the perceived

degree of difficulty or simplicity attributed to understanding and using cloud-computing

technology. Guided by a review of the literature related to the cloud-computing

ecosystem and the theoretical framework, the third hypothesis was framed theorizing that,

during the prediffusion stage, the lower the level small business leaders perceived the

complexity of cloud computing the greater their propensity to adopt the innovation. The

null hypothesis H30 was stipulated: No correlation exists between complexity and intent

to use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived complexity

of cloud-computing innovation. The scatterplot diagram depicting complexity and intent

to use, as shown in Figure 12, was presented revealing that the two variables were

linearly related such that as the attribute complexity increased the intent to use decreased

and inversely as the complexity attribute decreased the intent to use increased. The

regression equation for predicting the intent to use, given complexity, was

Predicted Intent to Use = -.25 Complexity + 6.86. (3)

An intent to use value of 6.86, which reflected the regression equation y-intercept value

when complexity theoretically equaled zero, was interpreted as a propensity to use

cloudcomputing technology between agree and strongly agree. The 95% confidence

135

interval for the slope, -.38 to -.13, did not contain the value zero, and therefore

complexity was significantly related to intent to use.

Figure 12. Scatterplot mapping complexity and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = -4.11, p < .01. The magnitude of the standardized

correlation coefficient was between medium and large, -.41, evidencing that complexity

was more than moderately related to intent to use. Approximately 17% of the variance of

the intent to use was associated with complexity. As hypothesized, the H30 null

hypothesis was rejected (p < .01), and complexity was negatively correlated to intent to

use. Therefore, based on these findings the third hypothesis, H3, was supported. The

conclusion related to the third research question was interpreted based on the statistical

136

analysis findings indicating that complexity was correlated to intent to use in a greater

than moderate inverse degree.

Research Question 4 Findings

The fourth research question asked: To what extent does observability, a perceived

attribute of innovation, relate to small business leaders’ intent to use cloudcomputing

services? The variable construct observability was operationally defined as the perceived

degree of visibility to others attributed to the results of the cloud-computing technology.

Guided by a review of the literature related to the cloud-computing ecosystem and the

theoretical framework, the fourth hypothesis was framed theorizing that, during the

prediffusion stage, the higher the level small business leaders perceived the observability

of cloud computing the greater their propensity to adopt the innovation. The null

hypothesis H40 was stipulated: No correlation exists between observability and intent to

use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived observability

of cloud-computing innovation. The scatterplot diagram depicting observability and

intent to use, as shown in Figure 13, was presented revealing that the two variables were

linearly related such that as the attribute observability increased the intent to use

increased. The regression equation for predicting the intent to use, given observability,

was

137

Predicted Intent to Use = .11 Observability + 5.83. (4)

An intent to use value of 5.83, which reflected the regression equation y-intercept value

when observability theoretically equaled zero, was interpreted as a propensity to use

cloud-computing technology nearing agree. The 95% confidence interval for the slope,

.02 to .23, did not contain the value zero, and therefore observability was significantly

related to intent to use.

Figure 13. Scatterplot mapping observability and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = 1.72, p < .01. The magnitude of the standardized

correlation coefficient was between small and medium, .19, evidencing that observability

was less than moderately related to intent to use. Approximately 3% of the variance of

138

the intent to use was associated with observability. As hypothesized, the H40 null

hypothesis was rejected (p < .01), and observability was positively correlated to intent to

use. Therefore, based on these findings the fourth hypothesis, H4, was supported. The

conclusion related to the fourth research question was interpreted based on the statistical

analysis findings indicating that observability was correlated to intent to use in a slightly

moderate positive degree.

Research Question 5 Findings

The fifth research question asked: To what extent does trialability, a perceived

attribute of innovation, relate to small business leaders’ intent to use cloud-computing

services? The variable construct trialability was operationally defined as the perceived

ability for use inspection attributed to cloud-computing technology. Guided by a review

of the literature related to the cloud-computing ecosystem and the theoretical framework,

the fifth hypothesis was framed theorizing that, during the prediffusion stage, the higher

the level small business leaders perceived the trialability of cloud computing the greater

their propensity to adopt the innovation. The null hypothesis H50 was stipulated: No

correlation exists between trialability and intent to use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived trialability of

cloud-computing innovation. The scatterplot diagram depicting trialability and intent to

use, as shown in Figure 14, was presented revealing that the two variables were linearly

related such that as the attribute trialability increased the intent to use increased. The

regression equation for predicting the intent to use, given trialability, was

139

Predicted Intent to Use = .12 Trialability + 5.63. (5)

An intent to use value of 5.63, which reflected the regression equation y-intercept value

when trialability theoretically equaled zero, was interpreted as a propensity to use

cloudcomputing technology between agree slightly and agree. The 95% confidence

interval for the slope, .04 to .21, did not contain the value zero, and therefore trialability

was significantly related to intent to use.

Figure 14. Scatterplot mapping trialability and intent to use coordinates. The coordinates,

best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = 2.78, p < .01. The magnitude of the standardized

correlation coefficient was medium, .29, evidencing that trialability was moderately

related to intent to use. Approximately 8% of the variance of the intent to use was

140

associated with trialability. As hypothesized, the H50 null hypothesis was rejected (p <

.01), and trialability was positively correlated to intent to use. Therefore, based on these

findings the fifth hypothesis, H5, was supported. The conclusion related to the fifth

research question was interpreted based on the statistical analysis findings indicating that

trialability was correlated to intent to use in a moderate positive degree.

Research Question 6 Findings

The sixth research question asked: To what extent does results demonstrable, a

perceived attribute of innovation, relate to small business leaders’ intent to use

cloudcomputing services? The variable construct results demonstrable was operationally

defined as the perceived degree to which tangible evidence of the benefits of a

cloudcomputing technology is communicable. Guided by a review of the literature

related to the cloud-computing ecosystem and the theoretical framework, the sixth

hypothesis was framed theorizing that, during the prediffusion stage, the higher the level

small business leaders perceived the results demonstrable of cloud computing the greater

their propensity to adopt the innovation. The null hypothesis H60 was stipulated: No

correlation exists between results demonstrable and intent to use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived results

demonstrable of cloud-computing innovation. The scatterplot diagram depicting results

demonstrable and intent to use, as shown in Figure 15, was presented revealing that the

two variables were linearly related such that as the attribute results demonstrable

141

increased the intent to use increased. The regression equation for predicting the intent to

use, given results demonstrable, was

Predicted Intent to Use = .23 Results Demonstrable + 4.71. (6) An

intent to use value of 4.71, which reflected the regression equation y-intercept value

when results demonstrable theoretically equaled zero, was interpreted as a propensity to

use cloud-computing technology between neither disagree nor agree and agree slightly.

The 95% confidence interval for the slope, .00 to .47, did not contain the value zero, and

therefore results demonstrable was significantly related to intent to use.

Figure 15. Scatterplot mapping results demonstrable and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

result was significant, t(84) = 2.01, p < .05. The magnitude of the standardized

142

correlation coefficient was between small and medium, .21, evidencing that results

demonstrable was less than moderately related to intent to use. Approximately 5% of the

variance of the intent to use was associated with results demonstrable. As hypothesized,

the H60 null hypothesis was rejected (p < .05), and results demonstrable was positively

correlated to intent to use. Therefore, based on these findings the sixth hypothesis, H6,

was supported. The conclusion related to the sixth research question was interpreted

based on the statistical analysis findings indicating that results demonstrable was

correlated to intent to use in a slightly moderate positive degree.

Research Question 7 Findings

The seventh research question asked: To what extent does voluntariness, a

perceived attribute of innovation, relate to small business leaders’ intent to use

cloudcomputing services? The variable construct voluntariness was operationally defined

as the perceived degree of self-determined sovereign choice to use cloud-computing

technology. Guided by a review of the literature related to the cloud-computing

ecosystem and the theoretical framework, the seventh hypothesis was framed theorizing

that, during the prediffusion stage, the more mandatory small business leaders perceived

the voluntariness of cloud computing the greater their propensity to adopt the innovation.

The null hypothesis H70 was stipulated: No correlation exists between voluntariness and

intent to use cloud computing.

A linear regression analysis was performed to evaluate the prediction of intent to

use cloud-computing technology based on small business leaders’ perceived voluntariness

of cloud-computing innovation. The scatterplot diagram depicting voluntariness and

143

intent to use, as shown in Figure 16, was presented revealing that the two variables were

not significantly related in a linear manner. The regression equation for predicting the

intent to use, given voluntariness, was

Predicted Intent to Use = -.01 Voluntariness + 6.22 (7)

An intent to use value of 6.22, which reflected the regression equation y-intercept value

when voluntariness theoretically equaled zero, was interpreted as a propensity to use

cloud-computing technology between agree and strongly agree. The 95% confidence

interval for the slope, -.27 to .25, did contain the value zero, and therefore voluntariness

was not significantly related to intent to use.

Figure 16. Scatterplot mapping voluntariness and intent to use coordinates. The

coordinates, best-fit line, and R2 coefficient were based on the normalized dataset.

The significance tests for the bivariate regression analysis included the t test

associated with the SPSS coefficients table output. The bivariate linear regression test

144

result was not significant, t(84) = -.07, p = .95. The magnitude of the standardized

correlation coefficient was negligibly small, -.01, evidencing that voluntariness was not

significantly related to intent to use. None of the variance of the intent to use was

associated with voluntariness. As hypothesized, the H70 null hypothesis was not rejected

(p = .95). Therefore, based on these findings the seventh hypothesis, H7, was not

supported. The conclusion related to the seventh research question was interpreted based

on the statistical analysis findings indicating that voluntariness was not significantly

correlated to intent to use.

Research Question 8 Findings

The eighth research question asked: Which has a stronger relationship to small

business leaders’ intent to use cloud-computing services the perceived attribute of

innovation relative advantage or voluntariness? Based on the theoretical framework, the

eighth hypothesis was conceived theorizing that, during the prediffusion stage, small

business leaders’ perceived the relative advantage of cloud computing was more strongly

related to their propensity to adopt the innovation than the strength of relationship

between their perceived voluntariness and propensity to adopt cloud computing. The null

hypothesis, H80, was stipulated: Voluntariness has an equal or greater correlation with

intent to use cloud computing than does relative advantage.

A multiple linear regression analysis was performed to evaluate the distinction

between small business leaders’ perceived cloud-computing attributes relative advantage

and voluntariness in predicting intent to use cloud-computing technology. The scatterplot

overlay, as shown in Figure 17, was presented revealing that the correlation between the

145

attribute relative advantage and intent to use depicted a positive linear slope while

depicting an insignificant linear slope for correlation between the attribute voluntariness

and intent to use. The multiple linear regression equation for predicting the intent to use,

given relative advantage and voluntariness, was

Predicted Intent to Use = .49 Relative Advantage + .04 Voluntariness + 2.97. (8)

An intent to use value of 2.97, which reflected the regression equation y-intercept value

when relative advantage and voluntariness theoretically equaled zero, was interpreted as a

propensity to use cloud-computing technology near disagree slightly. The 95%

confidence interval for the slope associated with predictor relative advantage, .26 to .72,

did not contain the value zero while the 95% confidence interval for the slope for

voluntariness, -.20 to .28, did contain the value zero. Therefore, based on the multiple

regression analysis 95% confidence intervals, relative advantage was deemed

significantly related to intent to use while voluntariness was not significantly related to

intent to use.

146

Figure 17. Scatterplot overlay mapping relative advantage and voluntariness with intent

to use coordinates. The coordinates, best-fit lines, and R2 coefficients were based on the

normalized dataset.

The significance tests for the multiple regression analysis included the t test for

each predictor with relative advantage was reported as t(83) = 4.27, p < .01 and

voluntariness was reported as t(83) = .34, p = .73. The magnitude of the standardized

correlation coefficients related to intent to use were reported as medium, .43, for relative

advantage while insignificantly small, .03 for voluntariness. As hypothesized, the H80

null hypothesis was rejected on the basis of the strength of correlation between relative

advantage and intent to use that was significant (p < .01) appearing greater than the

strength of correlation between voluntariness and intent to use that was not significant (p

147

= .73). Therefore, based on these findings the eighth hypothesis, H8, was supported. The

conclusion related to the eighth research question was interpreted based on the statistical

analysis findings indicating that relative advantage was correlated to intent to use in a

greater than moderate positive degree while voluntariness was not significantly correlated

to intent to use.

Summary Findings

Anchored in a prediffusion stage DOI theoretical framework, the overall intent of

this study was to examine the relationship between the perceived innovation attributes of

cloud-computing technology and small business leaders’ intent to use cloud computing.

The study intent was summarized by the overarching research question: What is the

propensity of small businesses to embrace the cloud-computing novel businesscomputing

paradigm? The analysis of the survey data collected for this study based on the

normalized dataset resulted in findings indicating significance in the relationship between

each cloud-computing attribute and intent to use cloud-computing technology with the

exception of the attribute voluntariness. The analysis findings were interpreted indicating

no significant relationship between the attribute voluntariness and intent to use cloud-

computing technology. The study findings are highlighted in Figure 18 by depicting

correlation coefficient values for each hypothesized predictor and criterion variable

relationship based on the predictive DOI theoretical framework applied to this inquiry.

148

Figure 18. Predictive DOI theoretical framework with hypotheses regression analysis

results. The theoretical framework conceptualization was designed to depict the predictor

and criterion variable relationship for each research question hypothesis. The correlation

coefficient (r2) is presented for each hypothesis indicating the strength of relationship

assessment.

The cloud-computing attributes relative advantage, compatibility, and complexity

were found to have a greater than moderate correlation to intent to use. These correlation

findings were consistent with other adoption inquiry electing to limit the measures of

innovation attributes to relative advantage, compatibility, and complexity because of their

strength of relationship with intent to use (Chong et al., 2009; Conrad, 2010; Rouibah &

Hamdy, 2009). In particular, prior research projected that the relative advantage of cloud

computing for small business was expected significant (Buyya et al., 2009; Etro,

2009; Gens et al., 2010; Greengard & Kshetri, 2010).

The cloud-computing attributes trialability, results demonstrable, and

observability were found to have a less than moderate correlation to intent to use while

149

voluntariness was found not significantly correlated to intent to use. These findings were

mostly congruent with Valier et al.’s (2008) open source software research results. Valier

et al. reported findings indicating that the relationships between each perceived

innovation attribute and use intent were statistically significant with the exception of the

correlation between voluntariness and use intent. The current study’s finding related to

trialability was consistent with Conrad’s (2010) DOI study of individuals’ willingness to

use personal web servers based on the factors relative advantage, complexity, and

trialability. Moreover, Conrad reported a positive correlation between complexity and

trialability, which suggested the opportunity for further analysis of this study’s data

beyond the initial hypothesis.

The cloud-computing innovation attributes most strongly related to small business

leaders’ intent to use the emerging technology included relative advantage, compatibility,

and complexity. The attributes relative advantage, compatibility, and complexity were

aligned more closely with the characteristics of the novel cloud-computing business

paradigm (Katzan, 2008, 2010; Mell & Grance, 2010; Truong 2010). Representing key

indicators for small business cloud-computing adoption, these findings were tied to

effective business practice as opportunity for economic value and competitive advantage

attributed to the cloud-computing business paradigm innovation, which included an

operationalized versus capitalized cost structure and an elastic global scale. Access to

capital financing and large-firm competition were portrayed in the research literature as

constraints to small business viability and economic contributions (Barringer & Ireland,

2010; Liao et al., 2008). Perceived competitive advantage improvements attributed from

150

the cloud-computing business paradigm were projected to be a significant factor in small

businesses use of cloud computing (Sharif, 2010; Truong, 2010).

The cloud-computing innovation attributes trialability, results demonstrable, and

observability were identified with the commercialization aspects of the cloud-computing

ecosystem. The extant research has shown cloud-computing technology was iteratively

re-inventive and not fully progressed to the commercialization stage of the

innovationdevelopment process (Etro, 2009; Rogers, 2003). Ambiguity related to the

cloudcomputing technology standards and commercialization maturity (Armbrust et al.,

2010; Blaskovich, & Mintchik, 2011; Buyya et al., 2009; Vouk, 2008) could have

factored into the less than moderate correlation between the perceived attributes

trialability, results demonstrable, and observability and small business leaders’ intent to

use cloud computing. The propensity to use and the adoption rate of IT innovation has

been shown to improve competitive advantage (Conrad, 2010; Truong, 2010). The

current study’s findings related to the cloud-computing attributes trialability, results

demonstrable, and observability were linked to effective business practice by advocating

opportunity for improved commercialization of cloud-computing technology.

A review of the literature (Bennett, 2009; Mell & Grance, 2009; Sharif 2010;

Smith, 2009; Wyld, 2009) indicated that policy and regulatory mandates have influenced

pre-commercialization adoption of cloud-computing technology by large companies and

federal government public sector agencies. The hypothesized relationship between

voluntariness and intent to use was not supported suggesting that the large company and

public sector cloud-computing mandates were not found to have a similar influence on

151

small business leaders’ propensity to adopt cloud computing. The perceived degree of

voluntariness appeared to have no significant correlation with small business leaders’

intent to use cloud computing, which was congruent with Valier et al.’s (2008)

prediffusion stage DOI research results related to open source software.

The conceptualization of voluntariness into the construct categories mandatory

and voluntary potentially introduced biases with respect to other research questions. A

determination of whether the study’s finding holds true when the data is segregated by the

voluntariness categories mandatory (voluntariness values less than four) or voluntary

(voluntariness values equal to or greater than four) was considered beyond the scope of

this research study. Follow-up post hoc analysis is recommended based on the

voluntariness categories mandatory and voluntary.

Based on the sampled small business leaders, on average, the study participants

indicated a negligible degree of trialability, moderately low degrees of complexity and

observability, and comparatively high degrees of relative advantage, compatibility, results

demonstrable, voluntariness, and intent to use cloud-computing technology. The findings

related to a high degree of intent to use differed from prior research (Buyya et al., 2009;

Etro, 2009; Gens et al., 2010) that suggested lower technology innovation adoption

tendencies by small businesses. However, this study’s findings related to a high degree of

intent to use was congruent with prior research that recognized small businesses were

sophisticated adopters of IT as a competitive tool (Chalhoub, 2010; Dibrell et al., 2008).

The low percentage of firms indicating extant use of cloudcomputing technology and the

152

types of cloud-computing technology used by these small businesses were consistent with

the findings of prior research (Schadler, 2009; Truong, 2010; Wyld, 2009).

The researcher acknowledged a number of inherent study limitations including the

convenience sampling based on the entire small business e-mail list maintained by ACA.

The study participants’ indication of a high degree of intent to use signaled the potential

for self-selection bias, which arises when individuals select themselves to participate in

research derived from nonprobability sampling (Heckman, 1979; Melino, 1982). A

related assumption included participant self-reporting bias, which occurs when a

participant’s experience, self-perception, and work environment influence their survey

responses (Han & Anantatmula, 2007). Valier et al.’s (2008) prediffusion stage DOI

probability sample research results were examined as a baseline (Blaskovich & Mintchik,

2011; Norton, Dow, & Do, 2008; Sampat, 2007) revealing a nearly identical high degree

of intent to use compared with this study’s results. The reviewer’s interpretation of the

analysis and findings presented in this research project were advised within the context of

these limitations.

This study’s conclusions and findings were framed as a basis for evidencing the

significance of this research project. The current study’s significance was demonstrated

by contributing to the academic literature via addressing the gap identified in extant

research concerning small businesses’ penchant to adopt cloud-computing innovation.

The significance of this study was demonstrated further by the prospect of applying the

study’s findings to professional practice as well as recognizing the research results’

potential implications for positive social change.

153

Applications to Professional Practice

The focal consequence of this research project resided in applicability of the study

findings to the professional practice of business. This study’s significance was realized

by identifying means for transforming the theoretical conclusions and findings into the

potential tangible benefits attributed via professional practitioner application. Numerous

opportunities to apply the study findings to professional practices were identified.

However, the application of the study findings was prioritized by emphasizing two facets

of small business economic vitality consisting of a cloud-computing deployment strategy

and a cloud-computing commercialization strategy.

Cloud-Computing Deployment Strategy

A holistic conceptualization of the novel cloud-computing business paradigm was

presented in this study including the synergistic integration IT components, IT capability

resourcefulness, and the operationalized capital-cost financial model. The

cloudcomputing ecosystem depicted in Figure 4 was presented to illustrate the technology

building-block components of cloud computing including IaaS, PaaS, SaaS, and other

technologies in conjunction with service delivery approaches such as private, community,

public, and hybrid clouds. The characteristics attributed to the cloud-computing

ecosystem highlighted IT capabilities such as on-demand self-service, ubiquitous access,

rapid elasticity, resource pooling, massive scale, metered services, and other IT facilities.

Moreover, the traditional front-loaded capital intense IT financial model was

reconstituted into an operationalized economic IT capability via the novel

cloudcomputing business model.

154

The attributes of cloud-computing innovation examined in this study via the

predictive DOI framework were measurements of the sample small business leaders’

perceptions related to the cloud-computing ecosystem. As depicted in Figure 18, this

study’s findings indicated significant (more than moderate) correlations between each of

the cloud-computing innovation attributes relative advantage, compatibility, and

complexity and the sample small business leaders’ intent to use cloud computing. The

significance of these correlation findings was accompanied by the sampled small business

leaders’ indication, on average, of a moderately low degree of complexity in conjunction

with relatively high degrees of relative advantage, compatibility, and intent to use cloud-

computing technology.

Notwithstanding the acknowledged limitations of this research project, this

study’s findings were deemed applicable to business by offering a lens for examining the

cloud-computing ecosystem for intended use. Examining cloud-computing technologies

via the cloud-computing attributes relative advantage, compatibility, and complexity were

projected to provide a viable framework for gauging small businesses’ use propensity.

This study’s findings were contemplated to inform small businesses formulation of cloud-

computing deployment strategies that leveraged its novel businesscomputing paradigm.

Moreover, these research results were expected to promote deployment of competitive

small business strategies by incorporating the potential benefits and capabilities attributed

to the cloud-computing ecosystem. The current study’s findings were recommended to

serve as a baseline for assessing the effectiveness of cloud-computing deployment

strategies for small businesses.

155

The findings from this study were interpreted as potentially signaling that the time

was opportune to shift the traditional IT strategy from a finite resource-based view to a

strategic perspective based on resourcefulness. Enabled by the operationalized pay-

foruse and global-scale IT capability of cloud-computing innovation, the resourcefulness

strategic perspective was envisaged to overcome the financial capital costs and

largecompany economy-of-scale advantage hindering small business economic viability.

Developing cloud-computing resourcefulness strategies was predicted to ignite small

business and entrepreneurial innovation via novel IT capabilities and disruptive creation

of products and service.

Cloud-Computer Commercialization Strategy

Strengthening the economic vitality of small business was shown to transcend to

economic community value while cultivating future returns via incubation of ongoing

entrepreneurial innovation. Given the dominating numbers of small businesses compared

to the number of large firms, the small business marketplace was forecasted as a

potentially large consumer of the cloud-computing ecosystem. Maturing the

cloudcomputing ecosystem beyond the prediffusion stage more fully into the

commercialization stage of the innovation-development process was anticipated to offer

significant economic contribution to and through small businesses. Cloud-computing

service providers and community policymakers were deemed to have vested interests in

promoting the commercialization of cloud computing for adoption by small businesses.

The cloud-computing attributes trialability, results demonstrable, and

observability were examined via the predictive DOI framework as means for assessing

156

the sample small business leaders’ perceptions in context of the commercialization

aspects of the cloud-computing ecosystem. As shown in Figure 18, this study’s findings

indicated significant (slightly moderate) correlations between the cloud-computing

attributes observability, results demonstrable, and trialability and the sampled small

business leaders’ intent to use cloud-computing technology. The significance of these

correlation findings was accompanied by the sampled small business leaders’ indication,

on average, of a moderately low degree of observability and a negligible degree of

trialability in conjunction with a relatively high degree of results demonstrable.

This study’s findings related to observability, results demonstrable, and trialability

were deemed applicable to cloud-computing service providers by informing their

commercialization strategies related to cloud-computing technologies. Recognizing the

potential for improving small business leaders’ perceptions related to the cloudcomputing

attributes observability and trialability offered opportunities for cloudcomputing service

providers. Specifically, cloud-computing service providers were advised to apply this

study’s findings in context of the predictive DOI framework by tailoring

commercialization strategies related to cloud-computing products and services to address

the needs and abilities of small business as opposed to large-firm cloudcomputing

offerings. This study’s findings were recommended to serve as a baseline for assessing

the effectiveness of cloud-computing service providers’ commercialization strategies for

small businesses. The findings from this study were interpreted as potentially signaling

that the time was opportune for cloud-computing service providers to capitalize on early

small business marketplace leadership.

157

Moreover, the study’s findings were anticipated to inform community, industry,

and academic policymakers as a motivation for formulating commercialization strategies

related to cloud computing. These strategies were conceived to entail targeted incentives

for collaborative transfer of cloud-computing academic research to private

entrepreneurial firms or targeted incentives for small business use of cloud-computing

services for innovation. Agencies similar to the ACA were anticipated to use the study

findings, in particular results demonstrable, to aid in collaborating with industry experts

to promote cloud-computing awareness in conjunction with education programs for

fostering economic development via small business sustainability and economic vitality.

This study’s findings were recommended to serve as a baseline for assessing the

effectiveness of community, industry, and academic policymakers’ commercialization

strategies for small businesses. Additionally, the application of the study findings to

professional practice served as the catalyst for implications related to social change.

Implications for Social Change

This research project’s implications for social change transcended small business

sustainability by contributing economic, societal, and environmental benefits. Founded

on the study’s significance, the implications for social change were derived from the

application of this study’s findings to professional practice, which emphasized strategies

for cloud computing deployment and commercialization. The focal implications for

social change were expressed in terms of tangible improvements related to workforce

preparedness, environmental consciousness, and social entrepreneur enablement.

158

Workforce Preparedness

Anchored in a predictive DOI framework, this study’s findings were instrumental

for framing the cloud-computing deployment strategy as a targeted application to

professional practice. An essential aspect of the cloud-computing deployment strategy

was expected to focus on development of personnel skills. This study’s findings related

to the perceived degree of difficulty or simplicity attributed to understanding and using

cloud-computing technology (complexity) were envisaged to inform small business

leaders about the personnel skill implications for people who support IT and people who

use IT to enact strategic outcomes.

These projected personnel skill implications were described in context of

workforce preparedness. In the small business context, IT personnel have engaged

tangibly in the deployment and support of the information systems technology. Requiring

less hands-on contact with the technology, cloud-computing IT skills were envisaged to

consist of virtual management and deployment of small business IT via the cloud-

computing ecosystem. The cloud-computing support skills were projected to require

greater governance and more tightly coupled integration with business strategy.

Similarly, small business executives were projected to adapt to new business acumen

associated with the cloud computing business paradigm. New skills were envisaged to

replace extant business strategies and performance metrics via operationalized proforma

and elastic global-scalable computing resources.

In essence, the societal implications related to small businesses were projected to

signal the obsolescence and transformation of designated personnel skills. The results

159

from this research project were planned to inform small business leaders’ forecast of

personnel skill requirements essential for sustainable advantage. While elevating

workforce competencies, developing personnel skills were projected to influence

reciprocally the perceptions about the cloud-computing attribute complexity for improved

use of cloud-computing technology.

Environmental Consciousness

Environmental concerns related to energy consumption and electronic-equipment

waste have grown to global proportions. Many regulatory agencies and nongovernmental

organizations (NGOs) have developed strategies to deal with these environmental

concerns. Transcending the economic benefits attributed to small business adoption of

the cloud-computing ecosystem, cloud-computing innovation was reported as consuming

approximately one third less energy and significantly reducing electronic equipment

disposal. In particular, the perceptions associated with betterment in lieu of precursor

ideas (relative advantage), results visibility to others (observability), and congruence with

experiences, values, and needs (compatibility) were envisioned to be inspired partially by

environmental consideration.

This study’s findings were projected to signal the potential for cloud-computing

deployment and commercialization strategies. The targeted cloud-computing strategies

were anticipated to have significant positive implications for small businesses by

decreasing energy consumption and reducing electronic equipment waste. Moreover,

these implications were envisaged potentially to enlighten policymakers regarding the

160

prioritization of regulatory and NGO environmental strategies for even greater

environmental outcomes.

Social Entrepreneur Enablement

Social entrepreneurs resemble business entrepreneurs in many tangible ways.

Business entrepreneurs have been characterized as innovating and incubating solutions to

industry-wide business problems. Similarly, social entrepreneurs have been recognized

for identifying and solving social problems of a grand scale. Throughout generations,

social entrepreneurs have engaged in societal callings. The societal results of social

entrepreneurship have reached global proportions because of the Internet, social media,

and the use of other information technologies.

Although social entrepreneurs typically are not profit motivated, the

operationalized IT capability projected via the novel cloud-computing business model

was envisaged to provide economic value to social entrepreneurs. The relative advantage

attributed to cloud-computing innovation ascribed to small businesses was thought

potentially to offer similar positive implications for social entrepreneurs. The perceptions

associated with betterment in lieu of precursor ideas, degree of difficulty or simplicity,

and congruence with experiences, values, and needs were considered similarly perceived

by social entrepreneurs. This study’s findings were considered applicable to social

entrepreneurs signaling the advantage of examining the transferability of small business

cloud-computing deployment strategies for the benefit of social entrepreneurs and other

nonprofit organizations.

161

Recommendations for Action

The recommendations for action presented in this study were developed as a

prioritized logical response to the application of the study findings to the professional

practice of business. The conclusions drawn from this study’s findings were reported

indicating significant correlations between perceived cloud-computing attributes and

small business leaders’ intent to use cloud-computing technology. This study’s findings

were interpreted as potentially signaling that the time was opportune for shifting the

traditional IT business strategy to the operationalized economic IT capability

characterizing cloud computing and improving cloud-computing commercialization

perspectives for small businesses. The primary audience summoned to action included

small business IT deployment decision-makers and cloud-computing commercialization

policymakers.

Although numerous recommendations were formulated as logical response to the

study conclusions, the proposals were culled to immediate prerequisite initiatives serving

as building-block steps to useful action. The call to action included three distinct

recommendations: dissemination of the study results, collaboration to formulate a small

business cloud-computing deployment framework, and a summons for further small

business cloud-computing research, which is discussed in the Recommendation for

Further Study section.

Diffusion of Study Results

The highest value related to study conclusions was anticipated to reside with the

small business IT deployment decision-makers and cloud-computing commercialization

162

policymakers. To promote optimal outcomes, dissemination of the study results to the

targeted audience was deemed essential. Based on DOI theory, viral communication

within members of the target audience was perceived the most effective means for the

dissemination of the study results. To facilitate this objective, prerequisite steps for

broadcasting the study findings were proposed including (a) publication of this doctoral

study via an online dissertation library for widespread circulation, (b) publication of a

summary report of the study results via Internet and e-mail distribution, and (c) in-person

presentation of the study results.

Publication of this doctoral study was planned using ProQuest as a means to

provide online access to a vast group of academic and practitioner researchers. In

satisfaction of the commitment to supply the study results to requesting survey

participants, a summary report highlighting the study results was planned. Distribution of

the summary report was planned via e-mail to study participants opting to receive the

study results. Additionally, access to the summary report of study results was intended as

a small business resource via ACA’s website as well as other websites.

By invitation, the researcher has committed to present the study results in person

to forums of small business leaders and entrepreneurs. The forums were planned as a

means for personal presentation of the study results envisioning interactive opportunities

for study results diffusion. These personal presentation settings were anticipated to

provide a channel for potentially launching small business cloud-computing deployment

collaboration.

163

Cloud-Computing Deployment Collaboration

A few small business recipients of the ACA announcement soliciting survey

participation have contacted the researcher requesting collaboration regarding the

development of cloud-computing deployment strategies. These small businesses have

assessed, in varying degrees of sophistication, the potential strategic benefit of

cloudcomputing technology for their respective firms. In response to the personal

presentations of the study results, additional small businesses were anticipated to join the

cloud-computing deployment collaboration. Additionally, cloud-computing service

providers were projected to participate in the small business cloud-computing

deployment collaboration.

The small business cloud-computing deployment collaboration was envisaged as a

focus group for constructing a small business cloud-computing deployment framework

based on the current study findings. As each small business has varying degrees of

capabilities, the development of the small business cloud-computing deployment

framework was projected to be an incremental and iterative process. Based on strategic

commitment and resource allocation, collaborating small business leaders were

anticipated to perform pilot testing of aspects of cloud-computing deployment framework

in their professional practice.

Recommendations for Further Study

Throughout the course of this research project opportunities for further study were

noted. Opportunities for additional research were identified in the study limitations,

application of the theoretical framework, literature review, and implication for social

164

change. Based on these observations, specific recommendations for further study are

proposed.

The convenience sample based on the entire small business e-mail list maintained

by ACA was identified as a potential study limitation. The generalization of the study’s

findings was limited potentially to the study population and was not implied beyond.

Replicating this study based on a randomized sample of a broader sampling frame was

recommended for further study. Moreover, comparing the results of this study against the

additional research conclusions and findings was also recommended.

As an alternative to the DOI framework, the systems of innovation approach was

recommended for understanding the use of cloud-computing technology by small

businesses. Applying the SIA framework was proposed for further study to explore

unconventional constructs related to economic, social, political, organizational,

institutional, and other factors that affect the diffusion and use of innovation. Comparing

the findings and conclusions of research based on the SIA framework against this study’s

results were thought to add a considerably broad understanding of small businesses’ use

of cloud-computing technology for economic vitality and sustainability.

The extant academic literature was virtually void of research related to small

business use of cloud computing to create and sustain their competitive advantage.

Building upon this study’s findings, further research is warranted to examine how

cloudcomputing deployment strategies can enable small businesses to create and sustain

their competitive advantage. This recommended inquiry was thought to offer significant

economic development opportunities to and through small businesses.

165

The preponderance of extant research about cloud computing was related to

largesize organization. The conclusions presented in this study addressed a gap in

understanding related to small business use of cloud computing. Although the mission

results perspective between social entrepreneurs and small businesses are typically polar

opposites, many of their strategies and key success factors are similar. To understand

more fully this study’s implications for social change, another recommendation for

further study was the transferability of this study’s conclusions and findings to social

entrepreneurs and nonprofit organizations.

Reflections

Acknowledging small businesses’ economic contributions, pursuing of a better

understanding of the early adoption of cloud computing by small business was rewarding.

Advancing the predictive DOI theory by examining the relationship between small

business leaders’ perceived cloud-computing attributes and their intent to use cloud

computing, this research project was designed with the expectation for application to

professional practice and the hope for positive societal implications. The rewards derived

from this research project were experienced beyond scholarly benefit by envisaging

practitioner value via application of this study’ findings for improved business outcomes

while recognizing positive implications for social change.

The researcher entered this project as a small business owner and entrepreneur

possessing more than 20 years of industry experience fostering IT agility and facilitating

strategically aligned IT innovations to achieve enterprise competitive advantage. A

conscious effort was made to ensure objectivity by mitigating the possibility of researcher

166

bias. Because of the nature of the quantitative cross-sectional survey strategy in

conjunction with strict adherence to ethical research practices, the researcher’s experience

was not perceived as posing a material bias to this study.

Although the researcher entered this project with a working understanding related

to the theoretical framework of DOI, a greater appreciation and understanding was gained

for the complexities associated with cloud-computing innovation attributes’ influence on

intent to use cloud computing in a prediffusion context. The researcher developed an

acute awareness of the importance and potential economic and societal value associated

with the research findings related to small businesses propensity to adopt cloudcomputing

technology. Moreover, application of the correlated study conclusions to professional

practice was projected possibly to signal a more rapid adoption rate of cloud computing

by small business, which potentially can improve the small business economic

contributions attributed to cloud computing by a factor of five times greater versus a slow

adoption rate.

Summary and Study Conclusions

Given small businesses’ economic and societal contributions, understanding small

business leaders’ propensity to adopt the novel cloud-computing business paradigm is

relevant in an era of economic turmoil and global competitiveness. A quantitative

research method using a nonexperimental cross-sectional survey design was employed to

examine the relationship between small business leaders’ perception of cloud-computing

innovation attributes and their propensity to adopt this emerging technology. A predictive

DOI framework using a pretested survey instrument was adopted to examine small

167

business leaders’ intent to use cloud-computing technology and their perceptions about

cloud-computing attributes including relative advantage, compatibility, complexity,

observability, trialability, results demonstrable, and voluntariness.

ACA cooperated in this research project by using their entire active e-mail list of

small businesses to solicit participation in this study’s online survey. Equaling slightly

less than a 4% response rate, 151 completed participant responses were collected via this

study’s web-hosted survey instrument. Based on the survey data collected, SPSS

software was used to perform descriptive statistics, reliability and validity analysis, and

statistical procedures for testing the hypotheses derived from this study’s eight research

questions. The results of the statistical analysis supported each of the hypotheses with the

exception of the seventh hypothesis associated with voluntariness.

Interpretation of this study’s findings indicated significant (more than moderate)

correlations between the cloud-computing attributes relative advantage, compatibility,

and complexity and intent to use cloud computing; significant (slightly moderate)

correlations between the cloud-computing attributes observability, results demonstrable,

and trialability and intent to use cloud computing; and no significant correlation between

the cloud-computing attribute voluntariness and intent to use cloud-computing

technology. The significance of these correlation findings was accompanied by the

sampled small business leaders’ mean indication of a negligible degree of trialability,

moderately low degrees of complexity and observability, and comparatively high degrees

of relative advantage, compatibility, results demonstrable, voluntariness, and intent to use

cloud-computing technology.

168

Notwithstanding the acknowledged limitations of this research project, the

conclusions drawn from this study’s findings were deemed to offer a lens for developing

strategies to promote rapid small business adoption of cloud-computing technology.

Examining cloud-computing technologies via the attributes relative advantage,

compatibility, and complexity were envisaged to provide a viable framework for gauging

small businesses’ use propensity while offering understanding needed to formulate

cloudcomputing deployment strategies that leverage its novel business-computing

paradigm. Evaluating cloud-computing technologies via the attributes observability,

results demonstrable, and trialability were envisaged to provide a viable framework for

gauging small businesses’ adoption propensity while offering understanding needed to

formulate cloud-computing commercialization strategies. The results from this research

project were anticipated to serve not only as a benchmark for assessing small businesses’

propensity to use cloud-computing technology but also as a catalyst for promoting rapid

adoption of cloud computing via deployment and commercialization strategies advocated.

This study’s conclusions and findings were framed as a basis for evidencing the

study’s significance. The study’s significance was partially demonstrated by contributing

to the academic literature by addressing the gap identified in extant research concerning

small businesses’ penchant to adopt cloud-computing innovation. The significance of this

study was further substantiated via the prospect of applying the findings from this

research project to professional practice by advocating small business cloud-computing

deployment and commercialization strategies. The study’s significance also was cast by

recognizing the potential implications for positive social change that were expressed in

169

terms of tangible improvements related to workforce preparedness, environmental

consciousness, and social entrepreneur enablement.

The findings from this research project were encouraging by offering a framework

for gauging small businesses’ penchant to use cloud-computing technology, which

potentially signaled a new era in small business sustainability and economic

contributions. Equally encouraging was the initial response from small businesses

desiring to collaborate regarding the development of a framework for cloud-computing

deployment strategies.

170

References

Al-Hajri, S., & Tatnall, A. (2008). Technological innovation and the adoption of Internet

banking in Oman. Electronic Journal of Organizational Virtualness & Networks,

10, 59-83. Available from http://www.ejov.org

Al-Qirim, N. (2007). The adoption of eCommerce communications and applications

technologies in small businesses in New Zealand. Electronic Commerce Research

and Applications: Intelligent Agents in E-Services, 6, 462-473. Retrieved from

http://www.elsevier.com/wps/find/journaldescription.cws_home/621289/descripti

on#description

Allen, K. R., & Stearns, T. M. (2009). Changing the equation for rural prosperity in a

global world [Abstract]. International Journal of Technology Transfer &

Commercialization, 8, 302-315. doi:10.1504/IJTTC.2009.024907

Anderson, P. L. (2009). The value of private businesses in the United States. Business

Economics, 44, 87-108. doi:10.1057/be.2009.4

Anderson, D. R., Sweeny, D. J., & Williams, T. A. (2008). Statistics for business and

economics (10th ed.). Mason, OH: Thompson South-Western Publishing.

Andrés, L., Cuberes, D., Diouf, M., & Serebrisky, T. (2010). The diffusion of the

Internet: A cross-country analysis. Telecommunications Policy, 34, 323-

340. doi:10.1016/j.telpol.2010.01.003

Archer, N., Wang, S., & Kang, C. (2008). Barriers to the adoption of online supply chain

solutions in small and medium enterprises. Supply Chain Management: An

International Journal, 13, 73-82. doi:10.1108/13598540810850337

171

Arizona Commerce Authority. (2010). Arizona Commerce Authority: Advancing

Arizona’s economy. Phoenix, AZ: Author. Available from

http://www.azcommerce.com/doclib/aca/FinalBrochure.pdf

Arizona Commerce Authority. (2011, February 25). Arizona innovation connection.

[Electronic mailing list message]. Available from

http://www.azcommerce.com/Home

Armbrust, M., Fox, A., Griffith, R., Joseph, A. D., Katz, R., Konwinski, A., . . . Zaharia,

M. (2010). A view of cloud computing. Association for Computing Machinery,

Communications of the ACM, 53(4), 50-58. doi:10.1145/1721654.1721672

Azadegan, A., & Teich, J. (2010). Effective benchmarking of innovation adoptions: A

theoretical framework for e-procurement technologies. Benchmarking, 17, 472-

490. doi:10.1108/14635771011060558

Barba-Sánchez, V., & Martínez-Ruiz, M. (2009). A longitudinal study to assess the most

influential entrepreneurial features on a new firm's growth. Journal of Small

Business and Entrepreneurship, 22, 253-266, 377. Available from

http://www.jsbe.com/

Bardhan, I. R. (2007). Toward a theory to study the use of collaborative product

commerce for product development. Information Technology and

Management, 8, 167-184. doi:10.1007/s10799-007-0013-y

Barringer, B. R., & Ireland, R. D. (2010). Entrepreneurship: Successfully launching new

ventures (3rd ed.). Upper Saddle River, NJ: Prentice-Hall.

172

Bates, T. (2005). Analysis of young, small firms that have closed: Delineating successful

from unsuccessful closures. Journal of Business Venturing, 20, 343-358.

doi:10.1016/j.jbusvent.2004.01.003

Benitez-Amado, J., Llorens-Montes, F. J., & Perez-Arostegui, M. N. (2010). Information

technology-enabled intrapreneurship culture and firm performance. Industrial

Management & Data Systems, 110, 550-566. doi:10.1108/02635571011039025

Bennett, J. (2009). Cloud computing: The future of internal comms? Strategic

Communication Management, 14(1), 9. Available from http://www.melcrum.com/

Blaskovich, J., & Mintchik, N. (2011). Information technology outsourcing: A taxonomy

of prior studies and directions for future research. Journal of Information Systems,

25(1), 1-36. doi:10.2308/jis.2011.25.1.1

Blau, J. (2009). New spurs for Europe's small firms. Research Technology Management,

52(3), 5-7. Retrieved from http://www.ingentaconnect.com/content/iri/rtm

Braganza, A., Awazu, Y., & Desouza, K. (2009). Sustaining innovation is challenge for

incumbents. Research Technology Management, 52(4), 46-56. Available from

http://www.ingentaconnect.com/content/iri/rtm

Braude, E. (2008). Software-as-a-service and offshoring. International Journal of

Business Insights and Transformation, 2(1), 93-95. Retrieved from

http://www.ijbit.org/

An Examination of Small Businesses Propensity to Adopt Cloud-Computing

Students also viewed