1.4.2.2 Methodological considerations in relation to the case studies and the meta-synthesis

Chapter 3 relies on the development of empirical research using qualitative case studies to examine how the element of participatory governance in tailor- made legal forms for social enterprises in the three selected jurisdictions works in practice. In Chapter 3, a fundamental methodological consideration regards the use of empirical research to examine the influence that tailor-made legal provisions concerning stakeholder participation exert on the governance of the social enterprise when they are implemented in practice. The combination of legal research with empirical research is common practice in legal studies. Many scholars agree that empirical legal research has substantially helped the study of the law and subsequently the development of the law.1

Another methodological consideration regards the enduring debate concerning the use of ‘qualitative versus quantitative’ methods in empirical research, which has also gained considerable impetus in empirical legal research.2 For instance, various scholars accept qualitative empirical evidence and its collection methods as a valid means of empirical legal research.3 Empirical legal scholars who are proponents of the qualitative research methods advocate that only qualitative research offers policy-makers valuable ‘information about decision- making, experiences and behaviour grounded in the experiences and world- view of those likely to be affected by a policy decision’.4 Others argue that empirical legal research only comprises research that makes use of quantitative data and methods of statistical analysis to identify patterns in judicial decisions.5 However, in general, legal scholars are encouraged to employ and develop various empirical methodologies in empirical legal research.6

What is more, a very important methodological consideration regards the validity and the reliability of the findings produced by the qualitative case studies. This methodological consideration is important due to: (i) the extant stereotypical perceptions against the use of case study in qualitative research, i.e. its perceived limitations and weaknesses in terms of validity and reliability;7 and (ii) the existence of methodological concepts developed in qualitative case study research to safeguard the validity, the rigour and the reliability of the case study findings.

One of the most serious and commonly claimed methodological concerns in the use of the qualitative case study method regards the lack of validity and objectivity in the sampling of the cases examined in the case studies.8 That is because the various logics applied to case study selection differ from the objective sampling logic applied to other empirical methods, e.g. surveys.9 It was earlier mentioned that the sampling (selection) logic that was applied to the selection of the case studies in Chapter 3 was developed on the basis of two concrete selection criteria, i.e. a purposeful and a theoretical sampling. The establishment of both criteria prevented a biased selection of cases, which is based on personal preferences of cases.

Another methodological concern in the development of the case studies regards the manipulation of data during the data collection stage and during their interpretation to such an extent that it reflects the personal bias of those conducting the research.10 This is a very serious methodological consideration, which imposes the use of various techniques during the case study development for the mitigation of the infiltration of any bias either in the data collection and/ or in the data analysis phase. In the development of the qualitative case studies although many interviews were conducted by me, a group of senior researchers and trained research assistants acted also as alternative and/or assisting data collectors in various occasions, e.g. at the interviews with Microstart, Volkshuisvesting, CORE and Breadshare.11 The assistance and alteration of data collectors limited the chances of having collected biased data, which are solely collected by me, i.e. by one single data collector. The same technique was applied in the coding of data where I was assisted by a group of research assistants and senior researchers who contributed as alternative coders.12 Additionally, for the mitigation of bias in the interview process, a similar basic questionnaire was used for each one of the four categories of interview respondents mentioned in the methodological Sub-sections of the case studies (see Annex I). The use of verbatim interview transcripts – instead of keeping interview notes in the data collection phase – also assisted in avoiding the infiltration of any personal bias during the interviews. Other techniques used to mitigate the bias during data collection were the technique of data triangulation and the cross-examination of data from different sources, and the validation of the data based on the received feedback from the interview participants. Those techniques prevented the existence of collected data, which are unreliable, inconsistent and subjective.

Another methodological consideration regards the validity and the value (i.e. usefulness) of the case study findings, which cannot be easily generalised, in particular when those are produced from individual cases.13 It was earlier mentioned in Sub-section 1.4.1 that the qualitative case studies are developed in multi-case study or single-case study settings. Each case study seeks the creation of contextual knowledge rather than creating generalisable knowledge. In addition, the case studies acknowledge their case-bound character in the relevant Sub-sections concerning ‘Limitation and future research’. In these Sub-sections, it is explained that the case studies’ objectives are to test the theory, in order to develop a generalisable theory, although generated from the cumulative examination of the selected variations of social enterprises. Accordingly, the development of the qualitative case studies is a process of accumulating case-specific knowledge from the different case studies that is ultimately aggregated and evaluated in the meta-synthesis.14

In addition, another methodological consideration regards the absence of control over the examined phenomenon – i.e. the implementation of participatory governance – and the absence of control particularly over the responses and behaviour of the interview respondents.15 Such a lack of control could have been resulted in the respondents altering their responses and/or their behaviour to the interviews because they know that they are being studied.16 Indeed, this important methodological consideration emerged particularly when one interview respondent in Case Study 8 required removing parts of his response in the interview transcripts with confidential information. However, the risk of losing data, which are perceived as confidential, classified or intimate, was not encountered in any other occasion. Nevertheless, this risk was always part of a trade-off for maintaining the naturalistic character of the case study research when examining the governance structure of social enterprises.17 In addition, this important methodological consideration also emerged in Case Study 9, which resulted in a limitation to the collection and analysis of interview data, i.e. the conduct of interviews with respondents from various organisational layers of the examined social enterprise. In Case Study 9, the young maturity of the selected social enterprise and the absence of its organisational structure allowed only a partial operationalisation of the protocol applied and replicated in the other 8 case studies. This inconsistency is discussed in detail in Sub-section 3.5.2.2 of Chapter 3. Accordingly, Case Study 9 was only partially considered in the meta-synthesis of the case study results as it is explained in the methodological Sub-section 3.6.2 of the meta-synthesis in Chapter 3.

Furthermore, the extant methodological concepts developed in qualitative case study research to safeguard the validity of findings are also considered in the development of the qualitative case studies, i.e. construct validity, internal validity, external validity and the parameter of reliability.18

Construct validity refers to safeguarding the process of operationalisation, i.e. identifying the proper operational measures for the examined concept.19 The methodological concept of construct validity aims to determine whether the case study contains any fallacies and whether it is developed in a way, which is free from bias by:20 (i) using multiple sources of evidence, which are crosschecked and cross-matched (i.e. the method of triangulation); (ii) establishing a chain of evidence in the presentation of the collected data, which anyone can trace; and by (iii) allowing key interviewees to review and provide feedback regarding the collected data.21

Internal validity regards the justification of the causal relationships in the provided inferences, which require the reconsideration of the analytical tools used by demonstrating the causal relationships between evidence and theory.22 Internal validity differs from external validity. External validity examines whether the results of a case study can be generalised and replicated, i.e. if the study will produce the same results should other scholars replicate it.23 To safeguard external validity in case study research various scholars suggest the ‘members’ checking’ technique mentioned above, during which interviewees can review the credibility of the collected data and their interpretations in the case study concerned, as well as the reliability and the replicability of the method used.24 A trail technique can be also applied according to which the research process and the outcome must be accessible, transparent, written down and catalogued.25

Acknowledging these methodological concepts of validity the methodological Sub-sections of the case studies demonstrate the use of multiple sources in the collection of data for the development of the case studies. Although as explained above, in one case study, i.e. Case Study 9, the collection of data was limited to collecting documents about the social enterprise rather than conducting interviews. The use of multiple sources is also demonstrated in the case studies’ data analysis Sub-sections named as ‘The case study results’ by reference to the different sources used for the development of the case study analysis. Accordingly, the chain of evidence is maintained across the developed case studies. However, as explained above, in one case study, i.e. Case Study 8, part of confidential data had to be removed, but this did not affect the outcome of the research.

I am also aware that the feedback and validation of the interview respondents could not be possibly demonstrated in the developed case studies. In that respect, references to follow-up interviews are provided in the case studies, whereas Figure 1.1 demonstrates a validated interview report by one interview respondent. In addition, Table 1.1 summarises all the measures used in the case studies to maintain the validity, credibility and reliability of the empirical research. The used measures are also explained in the methodological Sub- sections of the case studies.

Moreover, I bear in mind the sensitivity of collecting and analysing interview data and the ethical standards developed in qualitative research with respect to:26