Establishing quality criteria when conducting qualitative research is very crucial in achieving the validity and reliability of a qualitative study (Yin 1994). Healy and Perry (2000) established six criteria for evaluating the quality of qualitative research utilizing realism stance. These criteria are:
• Ontological appropriateness
• Contingent validity
• Multiple perceptions of individuals and or researchers
• Methodological trustworthiness
• Analytic generalization
• Construct validity.
Ontological Appropriateness refers to ensuring that the research problem deals with complex social science phenomenon that can be discovered through the re-flective perception of multiple participants even though it is only imperfectly un-derstood (Healy & Perry 2000). This requires selection of research problem that answers the how and why questions. As previously discussed, the research ques-tions answers the how and why firms opt for varying acquisition strategies at the time of market entry and changes in equity stake and post entry. More so, the con-text focus of the study and the process nature of acquisitions enhance the philo-sophical assumption of the study.
Contingent validity refers to validity about generative mechanisms and the con-texts that make them contingent (Healy & Perry 2000: 125). Thus, it implies the validity of the uncertainty about reality that may be contingently actualized. Hea-ly and Perry (2000: 125) suggest that contingent validity can be accomplished through theoretical sampling vis-à-vis literal replication and theoretical replica-tion. The aim of theoretical and literal replication is for confirmation of theory, extension of theory and improvement of the theoretical framework. Theoretical replication entails selecting cases that have different settings and are expected to accomplish different results (Yin 1994). Literal replication entails choosing cases that have similar settings and are expected to accomplish similar results (Yin 1994). In this study, this was accomplished via theoretical and literal replication.
As shown in Table 6 step 3, each additional cases was cautiously selected based on the theoretical framework (step 3b) and to replicate previous case(s) (step 3c).
Also, a case was selected that is polar opposite to produce opposing results but for predictable reasons (step 3d).
Furthermore, contingent validity was achieved by ensuring in depth questions were asked until the point of theoretical saturation. The context of each case was
described in terms of the appropriateness of the sources of the information, the size of the firms, industry, dates of the interviews and the managerial positions of the interviewee. Contingent validity was also achieved by ensuring that the re-sponding managers are the appropriate interviewee for the study. This was achieved by ensuring that only managers responsible for the acquisition decisions were interviewed. One could argue that why not interview the M&A team rather than just the acquisition managers? It is often that acquisition decisions involve M&A team who make target search and screening, due diligence, negotiation and the final decisions to acquire the targets. Nevertheless, the manager has the over-all picture of the strategic decision-making process. Thus, the managers provide strategic aspects that are nonetheless the focus of the current study.
Multiple perceptions of participants refer that in realism research, researchers should triangulate several data sources and several peer researchers’ interpreta-tions of those triangulainterpreta-tions (Healy & Perry 2000). Thus, the aim of multiple perceptions in realism research is similar to internal validity which aim at ensur-ing that researcher arrives at convergensur-ing lines of inquiry by data triangulation (Yin 2009: 115; Dooley 2002: 340). For the current study, this was achieved in four ways. First, through method triangulation, I ensured that the combination of interview data and secondary data sources generated consistency (Patton 2002).
For example, the acquisition strategies informed by the managers during the in-terview correspond to the strategies published on the company web pages and annual reports. Second, through triangulation of sources (several researchers’ in-terpretations of data triangulations); the interview data was given to a fellow re-searcher to ensure that the voice recorded version is same as the transcribed ver-sion of the interview (Patton 2002). Third, through analyst triangulation, the tran-scribed interviews were sent to the interviewee for verification (Patton 2002).
Fourth, the data was interpreted using multiple theories, referred to as theory tri-angulation (Patton 2002). Furthermore, as a way to provide the benefit of the research to the managers, it was promised that a summary of the findings and im-plication for managers will be sent to the managers. It is believed that the triangu-lation of data formats (audio and text) by managers and peer researchers provide a logical source of evidence (Yin 2009) and thus improves corroboration.
Healy and Perry (2000: 123) define methodological trustworthiness “as the extent to which the research can be audited by developing a case study database and by the use of quotations and matrices that summarizes key findings in the written report”. For the current study, methodological trustworthiness was achieved by the establishment of case study database. This is similar to the concept of reliabil-ity in positivist paradigm where methodological trustworthiness is established through ensuring that random error does not exist in a measurement. (ibid) The
reasoning behind the establishment of case study database is to ensure that the study can be audited. The case study database ensured the methodological trust-worthiness of this study. The case study database consists of three sections. Sec-tion (A) (Data CollecSec-tion Procedure) Consist of informed consent form (Appen-dix A) and interview guide questions (Appen(Appen-dix B) (Healy & Perry 2000: 122).
Section (B) (Data matrices-Table 8) consists of a summary of the data collected from each case study. Section (C) shows the interviewee anonymity and confiden-tiality confirmation (Appendix C).
Analytic generalization is one quality criteria in conducting a realism research. It ensures theory building rather than statistical generalization that is aimed at theo-ry testing (Healy & Pertheo-ry 2000). A realism research should aim at initially theotheo-ry building (i.e., conceptual development, operationalization, application), then con-firm or disconcon-firm the theory through the research data and then present a final theory or model or modified theoretical framework at the end of the report for future empirical testing or testing the generalizability of the final theory. In this study, research questions were identified in prior. This is followed by a literature review that leads to conceptual development (Section 4.4), application and opera-tionalization (Section 5). In chapter 6, the data collected was analyzed to confirm or disconfirm the theory. Chapter 7 follows a detailed discussion of the findings of the study as well as areas of future research.
Construct validity refers to how well constructs in the theory explored in the re-search are measured in the study under investigation (Healy & Perry 2000: 124).
Construct validity was ensured by using prior theory. The study under investiga-tion integrated two theories (instituinvestiga-tional theory and internainvestiga-tionalizainvestiga-tion process theory) that provided the constructs for this study. By utilizing these constructs, construct validity was ensured. The combination of case study database as de-scribed above and data triangulation discussed above enhances the construct va-lidity used for this study.
6 EMPIRICAL DATA AND RESULTS
This section discusses the empirical data and analysis as well as the results of the study. First, the empirical data or case study selected for this study is discussed.
Second, the background of case studies/firms is elaborated. Third, within cases analysis of the case study is presented. Fourth, cross-case analysis of the cases is done which presents the results of the study.