Source critical consideration

As with all research there are critical issues, for case studies some of the considerations that have to be attended to are confidentiality, anonymity and access according to Farquhar (2012). She refers to confidentiality as a general concern in regard to qualitative interviews due to the fact that respondents are sharing their personal opinions. This concern was reduced by making sure that all data was stored securely and by informing respondents of how their data was stored and used.

According to Farquhar (2012) anonymity is closely linked to confidentiality and refers to allowing

the respondents to remain anonymous in the study. In relation to this all respondents were informed that their participation was anonymous. Access refers to where and when the study was conducted and how that can limit the research depending on budget and selected respondents’ mobility (Farquhar, 2012). With this being a non-funded research study it was bound by budget to remain within a reasonable travel range, but since the research was conducted in a highly populated area this would not represent a significant issue in this study since.

Bryman and Bell (2011) also discuss ethics in business research by describing four general ethical principles that researchers need to be aware of; harm to respondents, lack of informed consent, invasion of privacy and deception. They describe harm to respondents as any harm that could come to the respondents of the study and most individuals view it as unacceptable. Harm could come to respondents in different shapes, it could be physical but it can also be less tangible as for example stress or risking employment in the future according to Bryman and Bell (2011). In accordance with this they recommend the same measures that are suggested by Farquhar (2012) in terms of confidentiality and anonymity. Bryman and Bell (2011) further relate the principle of lack of informed consent to providing respondents with enough information about the study to make a decision of whether or not to participate. This was done in this study by, prior to the interviews, presenting our general research area and motive for the study to possible respondents. The principle of invasion of privacy is described by Bryman and Bell (2011) as concerning the general right every individual has to one's own private opinions and thoughts. In this thesis this was relevant since respondents were asked to share private information about themselves in relation to the their role as a follower of an influencer. To ensure that there was no invasion of privacy the same measures as described above were taken in relation to anonymity and confidentiality. Respondents also had the right to refuse to any question they did not want to answer. The final principle, deception, is according to Bryman and Bell (2011) occurring in the event that researchers are not telling the truth to respondents about the purpose of the research study. To deal with this issue in this thesis it was clearly stated to respondents what the purpose of the study was.

During the interviews it was also found that the respondents sometimes struggled with explaining

their feelings especially when asked to explicate certain thoughts, beliefs and emotions. That may

have prevented access to possible unadulterated opinions. Lastly, there are also concerns associated

with language, since the interviews were held in English, which is not the native language of the respondents, some meaning could be lost.

In order to address quality, further the research design was reviewed with the help of four previously established research measures as suggested by Yin (2009). He describes how case studies can be tested with the help of four criteria commonly used to evaluate empirical social research; construct validity, internal validity, external validity and reliability.

5.8.1 Construct validity

Construct validity is most often connected with quantitative research and the search for accurate measurements of concepts (Bryman and Bell, 2011). Basically, the aim is that the study is investigating what it claims to investigate according to Farquhar (2012). Although this thesis has an interpretive standpoint and does not view reality as objective, construct validity has been addressed by following suggestions in Yin (2009). His suggestions relate to defining the research focus in terms of concepts as well as finding relevant measures for the research focus. In this thesis this was done with the help of the theoretical framework that delineates the aspects of the relationships between followers and influencers that is explored as well as by constructing an interview guide with open questions that relate back to the variables in the theoretical framework.

5.8.2 Internal validity

Internal validity relates to the notion that there are accurate links between the data and the results

(Bryman and Bell, 2011). Farquhar (2012) adapts this to a broader view by stating that the aim of

internal validity “is to persuade the reader that the research findings are based on critical

investigation of the data” (p. 101). According to Yin (2009) this mainly concerns the data analysis

stage of the study and Farquhar (2012) adds that internal validity can be achieved by providing

explanations for how the data was analysed. This was done in this thesis by carefully explaining

how the data was processed in terms of coding and both within-case analysis and cross-case

analysis. Since this is a qualitative study and Yin’s (2009) suggestions for internal validity are less

relevant for exploratory and descriptive research, this thesis adopts Kvale and Brinkmann’s (2009)

seven stages validation to add support for the understanding of internal validity. The seven stages

are not meant to inspect the research after completion but rather serve as quality control throughout

the thesis (Kvale and Brinkmann, 2009). The seven stages are thematizing, designing, interviewing,

transcribing, analyzing, validating and reporting. Thematizing relates to validity by developing

sound arguments and theoretical propositions, which in this thesis was done by thoroughly looking into previous research within the investigated topic and by building on theories used in prior similar projects. Designing involves validity through an able research designs, this was in this thesis done by adopting proper case study designs for investigating three similar cases. Interviewing relates to the validity of the interview responses in terms of quality and trustworthiness. According to Kvale and Brinkmann (2009), this can be achieved by asking the right questions during interviews such as

‘how’ and ‘why’, as well as understanding the respondents reasoning and the meaning of their answers. In this thesis this was assured by using an interview guide that allowed for comparison of interviews but it also made sure that the interviewers asked relevant open ended questions and planned prompts. The trustworthiness of the responses is crucial to the results of this thesis since the responses are the basis for the analysis. Thus it is important for the researcher to be aware of respondents’ social desirability meaning that they might respond in a certain way in order to be perceived in a good light (Bryman and Bell, 2011). Additionally, the stage of transcribing relates to producing valid transcripts of the interviews, in this thesis the informal tone of voice was kept throughout the transcripts to present answers as they were given and any notable emotion shown during the the interviews were also noted where emphasis on emotions were needed. Further, the stage of analyzing refers to the soundness of the interpretations, in this thesis this relates to how possible misinterpretations might arise from any unclear answers from respondents. To minimize this, this study sought to gain a brief understanding of the respondents through introductory biographical questions. The two final stages, validating and reporting relate to validating the findings in the right community and reporting the findings in a valid fashion. In this thesis, the findings were discussed among the researchers to settle on the adequate interpretation and all findings are reported both in Chapter 6 and full transcripts from all interviews are included in Appendix 2.

5.8.3 External validity

External validity is a general problem with qualitative research and relates to the generalizability of the findings (Bryman and Bell, 2011). Farquhar (2012) adopts this to the domain of case study research by describing different views on generalizing findings from case studies. According to her some researchers believe case study research cannot be generalized since the findings are unique to that specific case. She further describes how some researchers believe that multiple case studies are able to gain increased external validity by studying the same phenomena across several situations.

Yin (2009) contrast these opinions by stating that case studies are not aiming for statistical

generalization comparable to surveys, instead the aim is analytical generalization where the findings can be generalized to theory. In multiple case studies, Yin (2009) suggests that external validity can be increased through replication logic which means carefully selecting each case so that it either predicts similar results or contrasting results. This thesis attempts to increase external validity by, as has been previously stated, adopt literal replication as a replication logic, meaning that the cases studies are predicted to be similar. For this case the aim is thus to achieve theoretical generalization by proving that the theoretical framework is applicable in a larger extent. This can be supported if the connections described in the theoretical framework are exhibited similarly across the studied cases.

5.8.4 Reliability

Reliability relates to the degree by which the study can be replicated and might also be a general

issue with qualitative research according to Bryman and Bell (2011). They suggest that it might be

impossible to replicate a specific context since it can be highly situational in nature. Farquhar

(2012) broadens the context by noting that reliability in general refers to the stability and

consistency of the results. She further adds that in connection with replication another important

notion of reliability is transparency. Yin (2009) suggests that higher reliability can be achieved by

carefully documenting the process and steps of the study. An attempt to increase reliability in this

thesis is made by carefully documenting all the steps of the data collection and processing and by

being transparent by including transcripts of each interview.