Chapter 5 – Research Methodology
5.7 Data Collection Process
In this study, primary data was collected through interviews asking questions in uncontrolled situations which are used as the basis for twelve case studies. These interviews were divided into two categories: primary, with principal shareholder/owners and secondary, with other shareholders and/or advisors. Secondary datum such as business or exit plans and contracts were also sourced.
In this study multiple cases were examined for several reasons. Firstly, case studies are methodologically sound because they are based on the logic of replication (Parkhe 1993, Yin 2003a & 2003b). Secondly, the phenomena did not represent a rare occurrence or an instance of unusual access and therefore warranting a study of a single case. Thirdly, multiple cases are used to investigate complex behaviouristic phenomena and represent a greater variety of evidence (Yin 2003a). Finally, multiple cases offer the researcher an opportunity to triangulate evidence (Bonoma 1985, Eisenhardt 1989, Yin 2003a).
Miles and Huberman (1994) and Patton (2002) point out that there are no clear rules as to the correct number of cases. Eisenhardt (1989) recommends a range of four to ten cases as manageable, whereas Guba and Lincoln (1994) take the approach “as many as required to achieve theoretical saturation”. Perry (1998) believes that the acceptable range falls between two and four as a minimum, with an upper limit of fifteen. According to Patton (2002 p.244):
there are no rules for sample size in qualitative inquiry. Sample size depends on what you want to know, the purpose of the inquiry, what’s at stake, what will be useful, what will have credibility, and what can be done with available time and resources
Chapter 5 – RESEARCH METHODOLOGY
136
Keeping in mind that this is an exploratory study with the objective of proposing a new theory, theoretical saturation was the desired outcome which resulted in twelve cases being developed.
5.7.1
Case Selection
Purposeful or judgemental sampling techniques are typically used by qualitative researchers in the selection of case studies, as opposed to the systematic sampling techniques used in quantitative studies (Patton 2002). Researchers use their judgement to select cases where the phenomenon they are studying is most likely to occur (Neuman 2000, Collis & Hussey 2003). The logic and power of purposeful sampling is derived from the emphasis on in-depth understanding: “This leads to selecting information-rich cases for study in depth” (Patton 2002, p. 46). Patton further adds that studying information-rich cases produces insights and in-depth understanding, rather than empirical generalisations.
In this study, cases were selected on the following basis: 1. where a completed exit occurred, and
2. where the researcher was able to gain a level of rapport with the potential interviewee through personal contact or by introduction from a mutual contact.
In determining the number of interviews necessary to complete this study, two initial interviews per case were attempted. Firstly, the principal shareholder/owner was selected for the main interview (primary interview), with separate shareholder or advisors to the exit process for the supplementary interviews (secondary interview). Adding cases ceased when saturation of the concepts was achieved or when the researcher became convinced that the proposition was not going to be supported. However, due to single shareholdings
and unavailability of advisors in some cases, a total of only sixteen interviews were completed.
5.7.2
Interviews
Focussed interviews were determined as the most suitable format, and these were undertaken in semi-natural settings (Merton, Fishke & Kendall 1990). Yin (2003a) defines these as interviews which are open-ended and conversational in nature but following a certain line of questioning derived from the case study protocol and the interview guide. In this study the main interviews with owner/principal shareholders are referred to as primary interviews, and any subsequent interviews as secondary interviews.
Van der Heijden (1996) refers to interviewing from the perspective of scenario planning, and output from interviews can be viewed as scenarios. One of the main objectives of interviewing is to identify critical issues of concern. As it is important to make the interviewees feel that they are making a valuable contribution, the critical challenge is for the interviewer to be fully accepted by the interviewee and for there to be interaction between the two. Another important issue is an assurance that all information is kept confidential and for there to be no identifiable linkages with the data collected (refer Section 5.12 – Ethics).
Interviews can have shortcomings of being biased, having poor recall, and poor or inaccurate articulation (Yin 2003a). These weaknesses may stem from bias due to poorly constructed questions, or response bias, and reflexivity, where a respondent merely answers what they believe the interviewer wants to hear. In order to overcome these limitations, Yin (2003a) suggests three
Chapter 5 – RESEARCH METHODOLOGY
138
principles to be observed: use of multiple sources of evidence, creation of a case study database, and maintaining a chain of evidence.
Although the main source of data in this study was via primary interviews (Yin 2003a), where possible data collected was corroborated with other sources of information. The researcher undertook several strategies for corroborating data. Firstly, where secondary data in the form of printed Business Plans, Exit Plans, contracts of sale, reports, and Information Memorandums were available to the researcher, these were compared to the outputs from interview data. Yin (2003a) refers to this as non-converging evidence. In addition, data triangulation of a second source of interview data was established by interviewing a business partner or exit advisor for a specific case (Yin 2003a). This was then compared with data from primary interviews.
Following Yin’s (2003a) second principle for overcoming the limitations of interviews, all the collected case data in this study was organised into an evidentiary base consisting of case study notes, case documents (secondary data), and case narratives which were placed into separate storage folders. Finally, in accordance with Yin’s (2003a) third principle of ensuring robustness of the interview data; a chain of evidence was employed to allow retracing of steps leading from interview questions to logical conclusions.
To aid the interviews in this study an ‘evolved’ interview guide was developed. A series of questions was formulated from the Research Framework and Research Propositions (see Section 4.4.2) to probe issues during the interview process with the aim of answering the
research questions. The guide was designed to allow the researcher to accommodate real-life context and the fact that each case experience would vary markedly. The key purpose to this guide was to ensure that data collected during interviews remained relevant (Yin 2003a). This standardised interview guide was then revised following the pilot process.