Data Analysis Method

59 Reed. After terminating the audio recording, the participants are asked to give feedback on the interview and the clarity of all questions and topics covered. This is done in order to intensify the interviewer reflection aspect in depth interviews mentioned by Malhotra et al. (2017) and potentially improve flaws in the following interviews. In contrast to the expert interview, all consumer interviews are held in English, which facilitates their analysis and comparability. The consumer interview guide can be seen in Appendix 2.

60 complex because it is expressed through words. In the next step, data is summarized and classified into categories within the analysis by using conceptualization (Healey & Rawlinson, 1994).

In line with the above-discussed methods of data preparation for analysis, the collected data is examined.

Derived from the study’s abductive research approach, an abductive analysis approach of the data is applied, called abductive reasoning (Hansen, 2008). Within this creative approach of data analysis, theory is generated from qualitative observations. Hence, the first step of the analysis is to collect data and then explore the results by looking at which topic and issues to concentrate on (Corbin & Strauss, 2008).

Since all interviews are audio-recorded, the analysis starts with transcribing the audio files. To ensure the accuracy of the transcriptions, adequate software is used. The software Otter is applied for transcribing the English consumer interviews whereas Amberscript is applied for the German expert interview. Data cleaning is executed where language mistakes have obstructed the overall understanding.

Further, it is ensured that the transcription of the software was accurate. Where necessary, the transcription errors of the software are corrected, which equals a further form of data cleaning (Saunders et al., 2009). Subsequently, all 13 transcripts are read thoroughly. In order to analyze the transcripts of the interviews, two different methods are implemented. The transcripts of the expert interview and consumer interviews can be found in Appendix 3 and 4 respectively.

The expert interview is analyzed manually in an Excel sheet. More specifically, the transcript is summarized and categorized into the categories based on the expert interview guide. Finally, phrases and sentences belonging to the same topic are coded into text pieces and allocated to the specific category (Saunders et al., 2009). This analysis method is justified by the fact that only one expert interview is conducted. In contrast, the consumer interviews are analyzed with the software NVivo which facilitates the identification of patterns. Through the creation of nodes belonging to several categories, all consumer statements are coded and thus grouped for each node (see Appendix 5). This facilitates the overview as well as the analysis of the collected data. In order to enhance the analyzed patterns, relevant quotes of the consumers are highlighted and interpreted accordingly.

61 3.7.1. Credibility of Research Findings

A good research design is characterized by ensuring credible findings and is fundamental for every scientific study (Saunders et al., 2009) as it demonstrates that a genuine and accurate representation of the investigated topic is given (Shenton, 2004). With regard to the credibility of research findings, Guba (1981) has developed four quality criteria for trustworthiness within qualitative research. More precisely, the author states that credibility, transferability, dependability and confirmability are prerequisites for a trustworthy study.

Saunders et al. (2009) also determined the importance of the quality criteria of credibility in research findings. Accordingly, credibility is an indicator of the findings’ quality and is the first aspect which must be established and is linked to reliability and validity. While reliability specifies “the extent to which […] data collection techniques or analysis procedures will yield consistent findings” (Easterby-Smith, Thorpe, Jackson & Lowe, 2008, p.109), validity deals with the question whether the findings are really about what they appear to be about (Saunders et al. 2009). Further, validity provides insights about the relationship between two variables and whether it can be defined as a causal relationship.

Applied to the present study, a comprehensive knowledge on social media marketing, consumer engagement and brand equity is obtained through the literature review. This ensures that the research findings of the semi-structured interviews are credible. The acquired knowledge builds the foundation of the interview guides. Subsequently, the guides are used to structure the dialogues with the interviewees.

The thorough examination of all thematic areas before and during the interview enables the credible evaluation of the responses that constitute the qualitative data. Further, the credibility of the findings is increased as the respondents are informed about the topic prior of the interview, making it possible for them to engage with the subject.

The second quality criterion is defined as transferability of the research findings and discusses if the findings can be applied to other situations besides the specific research situation (Shenton, 2004).

However, this represents a challenge in qualitative research, as the sample sizes are rather small and findings dependent on the context of the study (Bryman & Bell, 2014). To address this difficulty, researchers are required to provide sufficient contextual depth of the study to facilitate transferability.

62 This is the prerequisite for determining if the findings can be legitimately transferred to other frameworks’ settings (Shenton, 2004).

This study utilizes an extensive description of the investigated topic as well as the research and data collection method in order to lead to a profound understanding, which represents a crucial prerequisite to make transferences to other studies (Shenton, 2004). By means of a detailed description of the illustrated company John Reed, all required information regarding the research are provided which facilitates the understanding and transfer of the results. With respect to the qualitative data collection method, i.e. interviews, information about the participants and the overall interview procedure is provided supporting the transferability of this study.

Dependability defines the third quality criterion and is a parameter of reliability that provides insights about the repeatability and consistency of a study’s findings (Bryman & Bell, 2014). Thereby, it is closely linked to the criterion of credibility as dependability ensures reliability to a certain extent (Shenton, 2004). Dependability can be generated if there is more than one researcher (Silverman, 2014). Therefore, conducting this study in pairs allows to analyze and compare the same data set from two separate perspectives ensuring dependability of the findings. Due to consulting an auditing approach, the study at hand provides findings (Bryman & Bell, 2014). In line with this approach is that all reports and documents of the whole research process, e.g. audio files, interview transcripts and data analysis documents, are saved to assure future research. Furthermore, the whole study process, including the research design and its implementation, is explained in detail to ensure the readers’ understanding and is complemented with the provision of an effective evaluation of the research. All these aspects contribute to the dependability of the findings (Shenton, 2004).

Lastly, confirmability builds the fourth quality criterion of research findings and is „the qualitative investigator’s comparable concern to objectivity” (Shenton, 2004, p.11). Confirmable findings demonstrate that the study has been carried out in good faith by the researcher (Bryman & Bell, 2014).

To achieve confirmability, researchers must implement actions to demonstrate that the findings emerge from the data and not their own predispositions (Shenton, 2004). More specifically, the findings have to

63 emphasize that the results are based on the respondents’ opinions are not the desired results of the researchers. Hence, the researcher’s bias must be completely overcome.

In this context, this study is providing confirmable results as it is characterized by a thorough explanation of as well as argumentation for every chosen method and approach. In addition, limitations of the study and further research approaches are illuminated, which also supports the reader’s understanding of every decision. Simultaneously, the reader can assess the extent to how far the data and its emerging framework may be accepted (Shenton, 2004).

3.7.2. Ethics of the Research Design

Researchers are confronted with ethical issues in every scientific research (Bryman & Bell, 2014). It is crucial to consider these issues when designing the research (Saunders et al., 2009) as they directly relate to the integrity of the research and the disciplines involved (Bryman & Bell, 2014). Thus, ethical issues implied by the research design must be carefully evaluated (Saunders et al., 2009). Therefore, Bryman and Bell (2014) identify eleven categories of ethical principles. Even though all categories must be considered when conducting a research, six, partly overlapping, categories were considered particularly important for the study at hand due to the qualitative data research approach. These categories are harm to participants, informed consent, privacy, confidentiality, anonymity and lastly, honesty and transparency.

Firstly, the category of harm to participants refers to “the potential to cause harm through the research process and the need to ensure physical and psychological well-being” (Bryman & Bell, 2014, p.71).

Even if there is a chance of harming the respondents before, during and after the research process, the well-being of the respondents is always the priority. This is interlinked to the criteria of informed consent, privacy, confidentiality and anonymity. Taking part in the research is voluntary and every participant is informed on the usage of data before the interview. Besides, all information regarding the participants as well as the gathered data itself is kept private and is treated strictly confidential. Lastly, every stage of the interview process is communicated to the participants openly and honestly which ensures the criteria of honesty and transparency. Conclusively, the well-being of the participants is respected at all times and ethical issues are treated responsibly.

64