Towards More Case Study Research in Entrepreneurship

Interest in entrepreneurship has been growing steadily in business schools, doctorate programs, and journal publications, raising the question of how research to advance the field is conducted. Theories on entrepreneurial processes and outcomes are drawn from a cross section of management science disciplines, including strategy, marketing, finance, operations, and organizational behaviour. If one considers entrepreneurship as encompassing vibrant processes of discovery, initiation, survival and growth, it is hard to imagine a research paradigm based predominantly on the hypothetico-deductive method. Case study research (CSR) provides an alternative method and opportunity to help build new theory across disciplines. It also offers an accepted framework for building theory from real-world, immediate observations rather than relying on retrospectives, surveys, or lab tests. Single-case studies, and longitudinal ones in particular, provide educators and practitioners with rich and valuable details from which to draw their own conclusions.


Introduction
Interest in entrepreneurship has been growing steadily in business schools, doctorate programs, and journal publications, raising the question of how research to advance the field is conducted.Theories on entrepreneurial processes and outcomes are drawn from a cross section of management science disciplines, including strategy, marketing, finance, operations, and organizational behaviour.If one considers entrepreneurship as encompassing vibrant processes of discovery, initiation, survival and growth, it is hard to imagine a research paradigm based predominantly on the hypothetico-deductive method.Case study research (CSR) provides an alternative method and opportunity to help build new theory across disciplines.It also offers an accepted framework for building theory from real-world, immediate observations rather than relying on retrospectives, surveys, or lab tests.Single-case studies, and longitudinal ones in particular, provide educators and practitioners with rich and valuable details from which to draw their own conclusions.
Entrepreneurship scholars such as Bygrave (2007;tinyurl.com/7vwkr49) have pointed to the decline in descriptive research and relevance to practitioners over the past decade, attributing the cause to an increasing difficulty in getting such studies published.Why is this?This article begins with a review of the CSR method as described by prevalent scholars, including a glossary of frequently encountered terms (Box 1).Using a survey of major journal publications over the past six years, the author explores how the method has been applied to entrepreneurship, and whether concerns over methodological rigour are justified.The author concludes with advice for researchers considering adopting the method.

What is Case Study Research?
CSR is a method whose defining features are: i) empirical study of contemporary situations in a natural setting; ii) a focus on asking "how" and "why" questions; and iii) the treatment of each case as an experiment in which the behaviors cannot be manipulated (Myers, 2009: tinyurl.com/7ts9bxe;Yin, 2009: tinyurl.com/7ywkcpy).Data collection can be accomplished using quantitative or qualitative methods, and it is a common misconception that case studies are based solely on the result of ethnographies or of participant observation (Dooley, 2002: tinyurl.com/7s5atpq;Yin, 2009).Case studies do not necessarily require fieldwork.In fact, they follow a linear, yet iterative process employing a variety of data collection methods to compare within and across cases for research validity.
Entrepreneurship as an emerging discipline has made good strides, but according to some, has fallen short of bringing its theory and literature up to the standards of others in the management sciences.Rich with the descriptive detail needed for insightful theory building in entrepreneurship, scholars have called for more case study research, particularly those incorporating non-retrospective and longitudinal observations.At the same time however, it has become rare to find such research published in A-level journals dedicated to entrepreneurship.A survey presented here of major entrepreneurship journals over the past six years revealed a publication rate of only 3% using the case study method.This presents a major impediment for developing fresh research in this field based upon the study of real cases.The author explores how the case study method has been applied to entrepreneurship research and provides recommendations for improved publication rates.
Discipline is the bridge between goals and accomplishment.

the criteria for interpreting the findings (if positivist)
In designing a case study, the researcher must initially establish that a case study is the best method to answer the research question at hand.The proposed case must be of sufficient interest to justify its production, because "a boring case is really just a waste of everyone's time" (Myers, 2009; tinyurl.com/7ts9bxe).Availability of secondary data and access to key informants of the case are essential and may limit both the number and selection of cases studied.Consideration of the initial constructs requires careful alignment of the propositions (2), units of analysis (3), and linking of the data (4), and is an essential activity of theory building (Eisenhardt, 1991;tinyurl.com/6wsppa4).The findings are often presented in an objectively narrative form, linking interview data with secondary data.
The selection of cases is chosen primarily to illuminate the research questions, and is generally based on theoretical replication, rather than sampling logic (Eisenhardt, 1991: tinyurl.com/6wsppa4;Yin, 2009: tinyurl.com/ 7ywkcpy).Multiple case studies are analogous to the replication of scientific experiments in order to determine whether a theory holds true under consistent circumstances (Yin, 2009).Depending upon the emergent theory being developed however, it may be advantageous to select cases based on contrasts or polar opposites, giving rise to "theoretical sampling".Eisenhardt (1989;tinyurl.com/7dfuc3z)suggests that between four and eight cases is optimal, although many examples of groundbreaking studies based on single cases exist.

Box 1. Glossary
Epistemological: The philosophical nature of knowledge; its presuppositions, extent, methods, and validity.
Hypothetico-deductive method: Method of scientific progress whereby a general hypothesis is tested by deducing predictions that may be experimentally tested; when falsified, a new hypothesis is required.
Longitudinal: A type of study that involves repeated observations of the same variables over long periods of time.

Participant observation:
A research strategy which aims to gain a close and intimate familiarity with a target study group though direct participation in their natural environment.
Positivist: A philosophical approach to knowledge based upon the deductive scientific method, i.e., development of testable, verifiable hypotheses with the goal of prediction, explanation, and theory development. www.timreview.ca

Criteria of Quality in Case Study Research
There are no commonly agreed upon quality standards in CSR, therefore one might ask: who is to be the judge of good quality?If the target audience is the academic community, then it is the inner circle of journal editors and reviewers, and theoretical, factual, or methodological insights are the underlying basis of assessment.If the gap between researchers and practitioners is to be bridged in order to advance the field, then the needs of the practitioner audience for more practical frameworks and suggestions needs to be considered.).Lincoln and Guba (1985; tinyurl.com/76xmfks) proposed, for example, that internal validity be recast as credibility: the degree to which a respondent's views fit with an inquirer's reconstruction; that external validity be viewed as transferability: the generation of sufficient case information so that case-to-case generalizations are possible; and reliability be considered as dependability: the degree to which the research process is logical, traceable, and documented.In qualitative research, these concepts, while more subjective, are alternative means of establishing the trustworthiness of the findings.
For case studies, Yin (2009; tinyurl.com/7ywkcpy)insists on strong triangulation of data sources (e.g., interview, observations, documents, archival records), to establish the reliability and validity of the research.Convincing the reader that adequate triangulation to support a finding has been achieved however, is an important yet subjective task left up to the researcher.

Use of the Case Study Research Method
Given that CSR is well suited for applied management fields such as entrepreneurship, it is surprising how little it is used as a method in entrepreneurship publications.Wigren (2007;tinyurl.com

Method Used in this Survey
There are currently two major journals among the Financial Times "top 45" (2010; tinyurl.com/88t68ep) that are dedicated to entrepreneurship: ETP and JBV.A review of articles from these two journals for the time period

Tom Duxbury
January 2005 to March 2011 was conducted to assess how successfully CSR has been applied to recent entrepreneurship research.

Sample selection
In the first stage of sample selection, the Scholar's Portal search engine (http://scholarsportal.info) was used to search titles, keywords, abstracts, and text bodies for the terms "case study", "case studies", or "case analysis".In the second stage, each of the previously identified articles was manually scanned to determine whether the article methodology meets the definition of CSR.A breakdown of the number of articles identified by journal is shown in Table 1.Table 2 provides a breakdown of the final number of articles selected by year of publication.

Review of articles and coding
Following selection of the cases, each case was reexamined in depth to determine: 1. how many cases were used in the study 2. the logic of case selection

Case attributes
Table 3 provides a synopsis of the case attributes, coded as previously described.The number of cases used in the studies examined ranged from one to 29, indicating that single-case studies do not appear to be a barrier to publication.The basis for case selection was split evenly between random selection, replication, variance and serendipity; two of the studies did not provide enough information to determine how cases were chosen.Most cases were conducted at a single point in time, and only 17% were conducted longitudinally, in line with Chandler and Lyon (2001;tinyurl.com/7pfgkwf).
The advancement of propositions in a study indicates a positivist application of the method.Half of the studies provided no theoretical propositions, which might be expected in a non-positivist epistemology.All case studies conducted interviews of some kind, and the majority were identified as semi-structured in the methods section.Somewhat surprisingly, 18% did not identify the type of interview, and only 17% used a CATA tool to assist in narrative analysis.
Triangulation of secondary data sources is essential to the CSR method, and 82% of the studies provided information regarding how this was accomplished.If one accepts Yin's (2009; tinyurl.com/7ywkcpy)positivist defini- Finally, most studies (56%) provided some evidence of validity, including: rechecking of interview data with informants, using independent coders, and reporting inter-rater reliability.

Level of analysis
Ten studies (55%) were single-level cases, in contrast to the 89% found by Chandler and Lyon (2001;tinyurl.com/7pfgkwf) in entrepreneurship research overall.Table 4 details the breakdown by level found in entrepreneurship CSR; the counts sum to more than the cases studied due to multiple-level studies.
It appears that the "firm" level receives 50% more attention than the "individual" level of analysis in this study and that this ratio has remained constant over the past decade, regardless of methodology.

Major topic areas of articles
The 18 articles were further coded into categories following entrepreneurship streams of research.Following Shane and Venkataraman's (2000; tinyurl.com/6sj7sk6)typology, the results are shown in Table 5.
It is apparent that, with nine studies (50%) focused on the "decision to exploit opportunities" stream, the field has been preoccupied with the individual entrepreneur in the recent past.

Discussion
This It would appear that many of the methodological criticisms of the entrepreneurship field highlighted by certain authors remain valid today.When we observe for example, that 18% of entrepreneurship CSR did not discuss their interview data collection method, or that a further 18% never triangulated their data, or that 44% never addressed methodological quality in any identifiable form, it may be true that we are "not there yet" (Crook et al., 2010;tinyurl.com/7fhuj37).
Regarding the second question, it is likely that CSR falls into the same challenges as other qualitative research methods: building new theory is difficult and requires more novelty, and there is a tension between breaking from existing theory while being attached to extant theory (Pratt, 2008;tinyurl.com/6tthf3f).Validation measures are unconventional and demand more of a trust relationship with the reader; consequently many researchers, reviewers, and editors prefer more positivist quantitative methods (Leitch et al., 2010;tinyurl.com/ 6qrvz8e).This may also be a reflection of the fact that: i) most major journals are based in North America and follow long traditions of positivist, quantitative research; ii) such research is well understood and requires less methodological reasoning; and iii) the entrepreneurship field comprises a relatively small number of researchers and journals compared to other management sciences.This lack of "critical mass" in the field generally, may limit the types of research we see undertaken and published.1. Always include secondary data sources and triangulation in the analysis.
4. Describe how CATA tools were used to improve consistency.
5. Always provide the reader with clues about reliability, validity, and trustworthiness.
In the end, CSR (along with all qualitative research methods) must excel at answering the basic reader questions of: "Is it an interesting story?", "Did it help me see organizational life differently?",and "Do I believe it?"(Pratt, 2008;tinyurl.com/6tthf3f).These are essential elements that will help CSR move forward as a research paradigm.

Advice for Researchers
Researchers considering using CSR in the field of entrepreneurship may expect to face many challenges, primarily: 1. Meeting methodological skepticism 2. Gaining access to entrepreneurial companies (some of whom fail, by definition) 3. Engaging in the type of longitudinal, multi-level studies being called for 4. Publishing work in the few A-level journals supporting entrepreneurship 5. Publishing work that is meaningful to the practice of entrepreneurship 6. Gaining tenure-track positions in a non-traditional management science discipline These are daunting propositions.It is recommended that the research design schema of Eisenhardt (1991; tinyurl.com/6wsppa4)and Yin (2009; tinyurl.com/7ywkcpy)be followed: build from theory, propose constructs that are operationally testable, and design the study from that foundation.This approach is well understood by editors, and a largely descriptive study is not likely to be worth the effort if publication is the end goal.Careful design of constructs and attention to data and theoretical triangulation is essential to meet expectations of rigour.
Many have pointed out that not everyone can perform qualitative research or execute case studies well.It is a craft work that would best be learned from a skilled craftsperson, rather than by texts alone.Simple publication of a case study article does not qualify it as an exemplar.New researchers are well advised to seek the mentorship of those experienced in conducting well-regarded case studies.

Conclusions
This study has confirmed the low publication rate (3%) of articles employing CSR in the two major entrepreneurship journals.It is unclear whether few are produced due to production challenges and low publication rates or whether many are produced and suffer high journal rejection rates.It is likely that the combination of these two factors has led to a circular publication impasse.This survey has found that, although general principles of CSR are followed by entrepreneurship researchers, there remains work to address the validity concerns voiced by editors.If publication is a goal, following well established paths laid out by Eisenhardt (1989;tinyurl.com/7dfuc3z) and Yin (2009;tinyurl.com/7ywkcpy) with substantial triangulation of data is essential.

Table 1 .
Number of articles identified in the first and second stages of this study, by journal 1. level of analysis 2. classification within major entrepreneurship themesChandler and Lyon's (2001; tinyurl.com/7pfgkwf)survey of entrepreneurship identified appropriate indications of validity (e.g., internal, external, construct) and reliability (e.g., inter-rater reliability, triangulation) that may be applied to CSR.

Table 2 .
Number of articles selected, by journal and year of publication www.timreview.caTowards

Table 3 .
Case attributes

Table 5 .
Entrepreneurship CSR streams of research www.timreview.ca