Constructing the study: research design and methods choices
4.5 Seeking qualitative rigour: reflecting on the research process
In doing qualitative research, it is important to discuss the validity and generalisability of the results in a greater context, in that propositions made are based on input from a limited, and selected, group of informants as opposed to a sample. The research presented in this PhD dissertation is no exception. In this section, I reflect upon the choice of qualitative methods for the research carried out and what implications it has for the general outcome.
As a researcher, it is important to state any potential biases that may stem from your relation to the object of study (Suddaby, 2006). I should therefore here state that
fashion is more than an object of study for me; in fact, it is one of my greatest passions in life. This of course brings with it both pros and cons in conducting research. While my wide interest for the subject matter can help in retracting information regarding the matters that are to be researched, it could be questioned whether or not I approach the research objectively, or if I rather have ulterior motives that would risk creating an interpretation bias of the data. For this reason, I believe the research design chosen to be highly suitable in that it clearly encourages as in- depth explanations as possible of all steps of the research process, including the development of the analysis through coding and clustering. The theoretical framing, using KIE as a concept to analyse fashion is also something I believe helps in avoiding any bias in that it forces me to rethink and re-evaluate phenomena I consider to know by heart, but which now need to be put into the context of entrepreneurship and knowledge usage/creation. The pre-knowledge has made it easier for me to acquire sufficient information about the sectoral innovation system that is fashion. Furthermore, my personal consumption of information around the fashion industry as a whole, including reading fashion press and more in-depth industry analyses, has throughout the process been channelled also to my research.
As highlighted in Section 4.1 above, Gioia et al. (2013) point out the importance of ensuring the critical view of what is being observed, as to not risk to fully adapt to what the informants say. In order to deal with this, it is suggested to have colleagues taking on the role of the devil’s advocate in relation to the results presented. Though I have worked on this project alone, as a PhD student, there is a continuous loop of feedback and critique from your supervisor(s) and fellow PhD students. This has been of great importance in making sure that the analysis performed has not become biased based on my own personal interest in fashion as a phenomenon. The continuous comments from people with less interest in, or experience of, fashion per se, and creative industries in general, have made me rethink my reasonings at times, thereby both sharpening the arguments, as well as avoiding biases based on personal opinions rather than actual research results.
As described in Section 4.3 above, the informants in the second round of interviews was not identical with those interviewed in the first round. The reason for this is that the same total group of individuals was contacted at both occasions, but different individuals responded to the request. It could be argued that this would mean that the material could not be used in its entirety to make an analysis of the effects that the two years that had passed had on the informants in relation to their potential desire of behaviour of becoming entrepreneurs. Seeing that the informants have the same background and furthermore that the interviews covered the same topics, the research can be considered valid. Furthermore, for the informants not included in the first round of interviews, basic questions to cover also their initial thoughts regarding entrepreneurship already at the point of graduation were included. As the most interesting data for the second round of interviews was that of what had happened in the last two years, that is the time from the first round of interviews onwards, I was
able to retract the same type of data from all informants. The study has further not included any type of intervention or test, meaning that all informants regardless of whether they were included in the first round or not still were able to give the same type of information.
All informants have the same academic background, meaning that the results presented essentially could be argued to only be valid for the specific setting of having obtained a BA and/or MA degree in fashion design from the Swedish School of Textiles. The fact that the informants for the two rounds of interviews were not identical could further be held forward as problematic in that the information regarding the initial intentions at the time of graduation could have been altered as they gained other experience over time. The same argument can be made with regards to the composition of the group of informants generally, as it contained two cohorts of graduates out of which one at the time of the first round of interviews already had one year of post-graduation experience. With regards to the second and third point made, two things should be noted. First of all, with regards to the group of informants not being identical in the rounds of interviews, a majority did participate in both rounds, creating a substantial overlap. Second, while it is true that all informants did not have the same amount of industry experience at the points of time for the interviews, the answers did not deviate and the conceptualisations are based on clear patterns that were valid for all the included informants.
In doing case study research, an important aspect is the question of saturation in terms of the data collected (Eisenhardt, 1989). No magic number in terms of informants exists for this, but rather it has got to do with the patterns that occur in the data as well as confirmation of already existing phenomena from the literature (Suddaby, 2006). To say that saturation has occurred is therefore difficult, and at the same time a necessity as to showcase that enough data to draw any conclusions has been collected. There is also the aspect of how many of the potential informants who actually agree to be part of the study. When analysing the coded material, clear reoccurring patterns with regards to the informants’ intentions and pathways could be observed. Therefore, theoretical saturation can be considered to have been achieved (Eisenhardt, 1989). To further test the validity of the results, a suggested next step for future research would be a more largescale study of fashion design graduates.
As education plays a central role in this research, it is also important to take into consideration any national peculiarities. In Sweden, all education, including at university level, is free, that is, the students have no tuition fees. In addition, all students are entitled to grants as well as beneficial, state-subsidised student loans for a total of six years of higher education. While students in general do graduate with minor debts due to their student loans, by international comparison, higher education in Sweden is very advantageous.
The designers studied in this research all have the same educational background, and as explained in more detail in Chapter 5, the Swedish School of Textiles in itself has a very specific structure to its programmes. To test the results’ generalisability, it would therefore be necessary to make comparative studies of graduates from other fashion schools. What this dissertation gives is that it proposes a framework to be further tested, in relation to fashion broadly, as well as other types of creative industries.
Through the interviews that form the basis of the research presented in this dissertation, I have been able to get qualitative understanding of fashion designers as potential entrepreneurs and the applicability of KIE as a theoretical concept in analysing the Swedish fashion industry. The data presented herein will give more in- depth discussions and ideas for further development in relation to the already existing literature, thereby contributing to the understanding of entrepreneurship in creative industries more generally.