EDMG611Wk4

The Qualitative Report 2018 Volume 23, Number 5, How To Article 1, 1043-1048

Collecting Sufficient Evidence When Conducting a Case Study

Pearl R. Smith University of Phoenix Online, Tempe, Arizona, USA

Case study is a popular research design within the social sciences despite

concerns of its credibility. Case studies provide an in-depth exploration of the

unit of analysis (case). Hence, data triangulation is a key characteristic of the

design whose purpose is to provide a thick, rich, and contextual description.

Data for varied sources enhances credibility of the study. However, studies

involving only one source of evidence exist in peer reviewed publications. This

paper reviews the nature of case studies and discusses the importance of data

triangulation. Further, three published case studies involving a single source of

data are reviewed and suggestions of more appropriate designs are provided.

Keywords: Case Study, Sources of Evidence, Interviews

Serving as the chair or member of dissertation committees, I frequently encounter

doctoral students who want to conduct a case study involving only interviews. When I inform

them case study entails data triangulation they often report seeing many published case studies

supported by only interviews and ask for permission to do likewise. When this occurs, the

mother in me wants to utter the age-old retort, “If you see everyone jumping off a bridge do

you want to follow them!” However, the researcher in me instructs them on the importance of

using a variety of data to enhance the credibility of their study, and I refer them authoritative

literature on case study design.

The popularity of case study design is growing in the social sciences (Thomas, 2011).

At the same time the credibility of the design is being questioned (Tight, 2010). Despite

concerns regarding credibility and limitations, researchers continue to gravitate toward the

design (Hyett, Kenney, & Dickson-Swift, 2014). Like my students, I also encounter many

published case studies built on a single type of data, and it troubles me. Equally disturbing are

studies in which the researcher purport using varied data sources in addition to interviews, but

neglects to fully describe the sources or analyze the data. Hence, a need exists to elaborate on

the importance of data triangulation in case studies to help promote proper use of the design.

This paper addresses the need for data triangulation in case studies, particularly as it

relates to qualitative studies. My purpose is not to make a case for using case study, but rather

to convey the importance of correctly employing the design. After discussing the nature of case

study design and the data triangulation requirement, I assess three case studies that involve a

single source of evidence and suggest alternative designs that may be more appropriate.

What is a Case Study?

Extant literature provides several definitions of case study. Stake (1995) defined it as

the study of an “integrated system” a “specific, complex, functioning thing” (p. 2). To Yin

(2009), it is a method used to understand a real-life phenomenon considering relevant

contextual conditions. Merriam (1998) wrote a qualitative case study is a particularistic,

descriptive, and heuristic study of a bounded system. The descriptive nature entails examining

the interaction of multiple variables presented via documents, quotes, samples, and artifacts. A

case can be a person, program, an organization, or a group (Yin, 2009).

Several factors must be considered when deciding whether case study is the appropriate

design. First the research question (Tetnowski, 2015). Yin (2009) explained a case study is

appropriate to answer “how” or “why” questions about a current event beyond the researcher’s

1044 The Qualitative Report 2018

control. A second factor is the specific design. Yin discussed three primary designs

(exploratory, explanatory, and descriptive. Stake (1995) offered two: intrinsic and

instrumental. Gerring (2004) emphasized, “case studies may be small- or large, qualitative or

quantitative, experimental or observational, synchronic or diachronic” (p. 353). Case studies

can involve one or multiple cases, requiring researchers to identify the suitable case(s) to study.

Another consideration is ability to obtain sufficient data. Despite its form, case studies

are data rich to enable a “thick description” (Hyett, Kenny, & Dickson-Swift, 2014) and in-

depth understanding (Creswell, 2013) of the case under investigation. Merriam (1998)

explained what makes a case study is the end, which should be a “holistic,” thickly descriptive,

and contextualized examination of the case. As such, a case study necessitates “deep and varied

sources of information” (Hancock & Algozzine, 2011, p. 16).

Data Sources

Yin (2009) espoused case studies rely on data triangulation and provided three

principles relative to data collection: (a) gather multiple sources of evidence to make findings

more convincing, (b) create a case study data base to house the raw data separate from the

research report, and (c) maintain a chain of evidence linking data from the research question to

the conclusions. In refuting assertions, case studies lack the rigor and objectivity of quantitative

studies. Flyvbjerg (2006) regarded the requirement to view the case through multiple lenses

strength of case study design. Such examination is required for the contextual understanding

necessary when examining the case (Cavaye, 1996). Snyder (2012) is an excellent example of

how the researcher’s need for data may grow during the data collection process. Hence, an

aspect of deciding if case study design is a fit is to ascertain whether sufficient data is obtainable

to support findings and conclusions.

Six sources of data typically are used in qualitative case studies: documentation,

archival records, interviews, direct observation, participant observation, and physical artifacts

(Stake, 1995; Yin, 2009). In general, interviews are the foundational data with other sources

used to validate information obtained from the interviewees (Hancock & Algozzine, 2011).

Hamilton and Corbett-Whittie (2013) advised using two or more sources or perspective that

will enhance understanding of the phenomenon. Although each type of data has strengths and

weaknesses (Hamilton & Corbett-Whittier, 2013; Hancock & Algozzine, 2011; Stake, 1995),

combined they are essential to the establishing the internal validity of a case study (Tetnowski,

2015).

Despite the requirement for data triangulation being prevalent in the literature (Gerring,

2004; Hamilton & Corbett-Whittier, 2013; Hancock & Algozzine, 2011; Stake, 1995; Yin,

2009) some researchers overlook this important aspect of conducting a case study, arguably

minimizing the quality of their work. I selected three single sourced case studies published in

peer reviewed publications to review. I had no involvement in these studies.

Case Analysis

Saldivar (2016) sought to identify factors causing Latino immigrants to leave or remain

in their places of employment in the California Central Valley to address four research

questions.

1. How do Latino immigrants in the California Central Valley understand and

describe the factors that cause them to remain at their places of

employment?

Pearl R. Smith 1045

2. How do Latino immigrants in the California Central Valley understand and

describe the factors that cause them to leave their places of employment?

3. How do current immigration and employment policies in the United States

affect a Latino immigrant’s employment status in the California Central

Valley?

4. How do Latino immigrants understand and describe their struggles or

obstacles they face in their search for employment in the California Central

Valley? (p. 12)

Data collected for the study comprised information from four focus group interviews.

Twenty-five workers participated, with four to 10 workers in each group. Participants included

permanents resident aliens and undocumented workers and represented two the agricultural

and manufacturing industries. However, the participants were not separated by these

distinctions. Such a separation would have enabled data analysis from multiple perspectives.

In addition, using focus groups may have influenced completeness or accuracy of information.

Saldivar reported some interviewees did not respond to questions, requiring him to draw

conclusions from reading their body language.

In addition, although the participants were given time off and “received support” from

their company to participate in the interviews, no data were gathered from the companies. Since

many of the interviewees’ comments dealt with policies and hiring practices, documents such

as company policy statements and hiring records may have been helpful to validate

interviewees’ claims. Interviews with supervisors and HR managers would also have provided

additional perspectives and much needed contextualization to fully understand the work

environment.

Two research questions guided Han, Kakabadse, and Kakabadse’s (2010) study which

explored the meaning of servant leader in among workers in China.

Would the concept of servant leadership have the same meaning in China, or, is

there an alternative term in the Chinese language that closely relates to the

concept of servant leadership? How do people construe the notion of servant

leadership in the public sector of the PRC? (p. 266)

Completing an opened ended questionnaire designed using the critical incident technique, 99

informants described workplace incidents they viewed as embodying servant leadership. The

researchers decided against using established case study tools of interviews and observations,

determining the survey alone was more culturally appropriate and would enhance participant

anonymity, honesty, and diversity of responses.

In this study designed to elicit workers’ perceptions, 85% of the respondents did not

provide stories of personal interactions with supervisors they felt reflected servant leadership

but rather provided abstract accounts. Because of this, the authors were unable to compare

dyadic leadership between managerial civil servants and non-managerial civil servants, which

Han et al. (2010) noted as a limitation of their study.

Armstrong (2015) studied how 25 project managers perceive the value of project

manager certification they had obtained. The researcher posed two research questions to

understand why the participants obtained the certification and how they felt about the decision.

1. Why do project managers who are volunteer participants of the research study

who are working as project manager in the new products and innovation area of

the consumer products industry decided to obtain a PMP Certification?

1046 The Qualitative Report 2018

2. Once project managers' volunteer participants who are working as project

manager in the new products and innovation area of the consumer products

industry obtained a PMP certification, did the research participant believe the

certification resulted in improved project management success; as evaluated by

the most recent project management success standards of cost, time, quality,

scope, stakeholder expectations and strategic alignment? (p. 30).

In addition to interviews, Armstrong stated using “researcher observations and any

other documentation that the case study participants were willing to share in support of the

research” (2015, p. 148). The researcher also purported maintaining a data base to ensure a

chain of evidence. However, data analysis entailed a thematic analysis based solely on the

interview responses. The researcher did not discuss what was observed, in what organization(s)

these observations occurred, what, if any, documents were provided; and how this data

supported accounts from the interviewees.

Agreement a case study entails data triangulation to promote contextual understanding

of the phenomenon is widely held. To this end researcher must have access into the

organization or environment where the phenomenon is occurring (Hancock & Algozzine,

2011) to obtain needed evidence. When these conditions cannot be met researchers should

consider other qualitative designs.

Possible Alternative Designs

A case study is ideal for studying a process Merriam (1998) and for generating

descriptive inferences, which is important in social science research (Gerring, 2004). However,

other designs are more appropriate for studying participants’ perceptions. For example,

narrative inquiry may have been a better design for Saldivar (2016) whose research questions

focused identifying factors causing the participants to stay on or leave a job and the obstacles

they faced finding employment. One could argue these are separate topics, making the study

lack boundedness, which is a characteristic of case study (Merriam, 1998; Stake, 1995; Yin,

2009). In narrative studies participants share stories depicting their view of the world (Connelly

& Clandinin, 1990). Using this design would likely enhance participation because participants

could answer questions verbally or in writing if they found doing so uncomfortable. In addition,

although multiple tools can be used to gather data, triangulation is not required. Finally,

narrative inquiry engenders stories solely reflecting the participants’ perspective, without

regard of context.

Another design that may have been a better fit for Saldivar (2016) is phenomenology.

The design, which involves one-one-one interviews, enables a study of appearances through

participants’ perspective taking (Moran & Mooney, 2002). Like narrative inquiry, there is no

requirement to contextualize or triangulate the data because the intention is to draw conclusions

and themes from participants’ perceptions. To Husserl, the father of phenomenology,

perception alone is evidence of one’s experience (Moran & Mooney, 2002).

Similarly, narrative inquiry or phenomenology may be a better fit for Armstrong (2015)

whose interview-based study lacked an “appreciation of context” characteristic of case studies

(McLeod & Elliot, 2011). As stated, narrative inquiry is an exploration of a person’s lifeworld

through participant storytelling. Moustakas (1994) explained phenomenology focuses on

participants’ reflections, and “is concerned with ideas and essences” (p. 5). That design may

align with Armstrong’s goal to explore the participants’ perceptions and feelings about

obtaining project management certification.

Given the political and social conditions in China, Han et al.’s (2010) use of a survey

questionnaire was prudent to minimize risk to the participants. However, calling the study a

Pearl R. Smith 1047

case study is a misnomer because it lacked context and varied data sources. Yin (2009) stressed

an in-depth understanding of a phenomenon requires considering important contextual

conditions, and surveys alone cannot provide contextual understanding. Han et al. (2010) is

better described as a survey study.

Narrative inquiry and phenomenology both have nuances researchers must consider in

their design selection. However, neither requires data triangulation associated with case study.

Further, Yin (2009) explained a survey questionnaire is a single data source, and a survey is a

distinct form of qualitative inquiry.

Conclusion

Although critics question the credibility of case studies, the design still has wide appeal

for social science researchers. Case studies are data rich, contextual, and involve multiple

sources of evidence. Advocates of the design acknowledge analyzing so much qualitative data

is laborious (Cavaye, 1996; Hancock & Algozzine, 2011). However, data triangulation is

widely regarded as a means to enhance the credibility of case studies. When researchers are

limited to obtaining a single source of data choosing another qualitative design is

recommended.

References

Armstrong, A. L. (2015). The value of project management certification in achieving project

success: A qualitative study (Doctoral thesis). Retrieved from ProQuest Dissertations

& Theses Global. (Order No. 3714323)

Cavaye, A. L. M. (1996). Case study research: A multi-faceted research approach for IS.

Information Systems Journal, 6(3), 227-242.

Connelly, M., & Clandinin, J. (1990). Stories of experience and narrative inquiry. Educational

Researcher, 19(5), 2-14. Retrieved from http://www.jstor.org/stable/1176100

Creswell, J. W. (2013). Qualitative inquiry & research design: Choosing among five

approaches. Los Angeles, CA: SAGE.

Flyvbjerg, B. (2006). Five misunderstandings about case-study research. Qualitative Inquiry,

12(2), 219-245. doi: 10.1177/1077800405284363

Gerring, J. (2004). What is a case study and what is it good for? The American Political Science

Review, 98(2), 341-354. Retrieved from http://www.jstor.org/stable/4145316

Hamilton, L., & Corbett-Whittier, C. (2013). Carrying out your case study. In L. Hamilton &

C. Corbett-Whittier (Eds.), Research methods in education: Using case study in

education research (pp. 83-116). London, UK: Sage. doi: 10.4135/9781473913851.n6

Han, Y., Kakabadse, N. K., & Kakabadse, A. (2010). Servant leadership in the People's

Republic of China: A case study of the public sector. The Journal of Management

Development, 29(3), 265-281. http://dx.doi.org/10.1108/02621711011025786

Hancock, D., & Algozzine, B. (2011). Doing case study research: A practical guide (2nd ed.).

New York, NY: Teacher College Press.

Hyett, N., Kenny, A., & Dickson-Swift, V. (2014). Methodology or method? A critical review

of qualitative case study reports. International Journal of Qualitative Studies on Health

and Well-being, 9(1), 23606. https://doi.org/10.3402/qhw.v9.23606

McLeod, J., & Elliott, R. (2011). Systematic case study research: A practice-oriented

introduction to building an evidence base for counselling and psychotherapy.

Counselling and Psychotherapy Research, 11(1), 1-10.

https://doi.org/10.1080/14733145.2011.548954

Merriam, S. B. (1998). Qualitative research and case study applications in education: Revised

1048 The Qualitative Report 2018

and expanded from Case Study Research in Education. San Francisco, CA: Jossey Bass.

Moran, D., & Mooney, T. (Eds.). (2002). The phenomenology reader. New York, NY:

Routledge.

Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.

Saldivar, A. P. (2016). A case study of Latino immigrant employment in the California Central

Valley (Doctoral dissertation). Retrieved from ProQuest Dissertations & Theses Global.

(Order No. 10042680)

Snyder, C. (2012). A case study of a case study: Analysis of a robust qualitative research

methodology. The Qualitative Report, 17(13), 1-21. Retrieved from

http://nsuworks.nova.edu/tqr/vol17/iss13/2/

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Tetnowski, J. (2015). Qualitative case study research design. Perspectives on Fluency and

Fluency Disorders, 25(1), 39-45. doi:10.1044/ffd25.1.39

Thomas, G. (2011). A typology for the case study in social science following a review of

definition, discourse, and structure. Qualitative Inquiry, 17(6), 511-521,

doi:10.1177/1077800411409884

Tight, M. (2010). The curious case of case study: A viewpoint. International Journal of Social

Research Methodology, 13(4), 329-339. doi:10.1080/13645570903187181

Yin, R. (2009). Case study research: Design and methods (4th ed.). Thousand Oaks, CA: Sage.

Author Note

Pearl R. Smith, PhD is an adjunct professor and operates an independent training and

development practice specializing in leadership development. She currently teaches in the

organizational leadership program and University of Phoenix and Mercer University. Her

clients include many state and federal government agencies. She is a federal government retiree

who served in numerous staff and management positions prior to establishing her private

practices. Dr. Smith holds a PhD in Organizational Leadership from Regent University.

Correspondence regarding this article can be addressed directly to: [email protected].

Copyright 2018: Pearl R. Smith and Nova Southeastern University.

Article Citation

Smith, P. R. (2018). Collecting sufficient evidence when conducting a case study. The

Qualitative Report, 23(5), 1054-1048. Retrieved from

https://nsuworks.nova.edu/tqr/vol23/iss5/2

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.