1. Introduction
4.7 Data analysis
4.7.2 Coding process
As referred to in chapter 4.4, coding serves as the underlying tool to carry out the analysis and presents an essential processes within GT (Bryman, 2012). Coding in the context of qualitative data analysis entails “labelling each unit of data (…) with a code that symboli[z]es or summary[z]es that extract’s meaning” (Saunders et al., 2016, p. 580). The coding procedure addresses the non-standardized form of the qualitative data and serves in order to classify it into categories. Thereby the qualitative data becomes more structured and manageable (Saunders et al., 2016). Aligned with the GT, which serves as both strategy and method, it presents an iterative process with the aim of discovering relevant themes, and subsequently developing a theory out of the categorized of data. The relevance of each code can, inter alia, be justified by its reoccurrence (Strauss & Corbin, 1998).
The codes are thereby created in vivo and not with a clearly defined framework or hypothesis back-ground based on already existing theory. Thereby, theory is developed throughout the process of the data collection and -analysis. It is aimed to gather the meanings out of the perception and social contexts of the participants in the context of the research question and to correspondingly draw theo-retical conclusions (Saunders et al., 2016). Aligned with the GT key element of constant comparison, the codes are repeatedly revised and adapted in the light of the added interviews and throughout the course of the analysis process (Bryman, 2012).
In regard to the research at hand, the coding process is oriented on the approach of Strauss and Corbin (1998). The coding process is differentiated in three parts, which lastly yield in grounded theories (see figure 7). Each of the three stages – open, axial, and selective coding – is elaborated on in more detail within the following chapters.
It should be noted that the coding process is carried out by both of the researchers and with the help of the software NVivo.
Figure 7: Coding process according to Corbin & Strauss (1998) (adapted from Saunders et al., 2016)
4.7.2.1 First stage: Open coding
Following the approach of Strauss and Corbin (1998), the coding process starts with the first stage, namely open coding. As such, one understands “the disaggregation of data into units” (Saunders et al., 2016, p. 595), which entails that text passages in the transcripts, that are considered relevant for the research question in the widest sense, are assigned a label that roughly summarizes the content of the text passage. In this manner, the entire data is covered. Following the GT’s key element of con-stant comparison, it must be thereby ensured that data sets with very similar contents are given the same label during the coding process. Throughout the entire coding process, emphasis is put on the derivation of meaning, which is deemed the overall objective (Saunders et al., 2016; Strauss &
Corbin, 1998). As a result of such open coding, one achieves “the creation of a multitude of concep-tual labels related to the lower level of focus and structure” (Saunders et al., 2016, p. 596).
In regard to the investigation of the different types of CE within the luxury fashion industry and the influence of CE with LFBs on the customers’ SC, a large number of frequently occurring concepts were initially noted in the so-called nodes. Thereby, it is first of all attempted to include all relevant phenomena that can be broadly linked to answering the research question; it is aimed to not be biased by knowledge based on prior theory.
The following table shows an exemplarily overview of two open codes and the quotes they are based on (see table 6). However, the representation in the table below is not complete – in terms of quotes and the amount of codes – and serves only as an example to illustrate the coding process. A complete version of the list of codes can be found in the appendix (see appendix C).
4.7.2.1 First stage: open coding
Following the approach of Strauss and Corbin (1998), the coding process starts with the first stage, namely open coding. As such, one understands “the disaggregation of data into units” (Saunders et al., 2016, p. 595), which entails that text passages in the transcripts, that are considered relevant for the research question in the widest sense, are assigned a label that roughly summarizes the content of the text passage. In this manner, the entire data is covered. Following the GT’s key element of con-stant comparison, it must be thereby ensured that data sets with very similar contents are given the same label during the coding process. Throughout the entire coding process, emphasis is put on the derivation of meaning, which is deemed the overall objective (Saunders et al., 2016; Strauss &
Corbin, 1998). As a result of such open coding, one achieves “the creation of a multitude of concep-tual labels related to the lower level of focus and structure” (Saunders et al., 2016, p. 596).
In regard to the investigation of the different types of CE within the luxury fashion industry and the influence of CE with LFBs on the customers’ SC, a large number of frequently occurring concepts were initially noted in the so-called nodes. Thereby it was first of all attempted to include all relevant phenomena that could be broadly linked to answering the research question; it was aimed to not be biased by knowledge based on prior theory.
The following table shows an exemplarily overview of two open codes and the quotes they are based on (see table x). However, the representation in the table below is not complete – in terms of quotes and the amount of codes – and serves only as an example to illustrate the coding process. A complete version of the list of codes can be found in the appendix (see appendix x).
OPEN CODE RESPECTIVE QUOTE(S)
“Well, first off, I get the brand, the anomaly that this kid from a low-income suburb in the Philippines is wearing it. There’s that number one. Because people like me are not sup-posed to be wearing these things or can purchase these things. So that’s cool. I call that swag.” (Int. 3)
“I think, I like of course the quality of the product material, the product itself, but also
4.7.2.2 Second stage: Axial coding
Axial coding is defined as the second coding stage subsequent to open coding. As such, one under-stands “a set of procedures whereby data are put back together in new ways after open coding, by making connections between categories” (Strauss & Corbin, 1990, p. 96). This second step entails to link codes “to contexts, to consequences, to patterns of interaction, and to causes” (Bryman, 2012, p.
578). It is aimed to identify relationships which can be drawn between the several open codes. By doing so, axial coding can be described as the second level working towards the process of theoretical development. In this step, such relationships, however, are not only identified but furthermore rear-ranged in order to get a hierarchical order between the different codes, so that sub-categories emerge.
Those, in turn, need to be verified against the conducted data (Saunders et al., 2016).
Again, an exemplarily overview of the axial coding is illustrated in the table below (see table 7).
Table 7: Exemplarily axial coding
4.7.2.2 Second stage: axial coding
Axial coding is defined as the second coding stage subsequent to open coding. As such, one under-stands “a set of procedures whereby data are put back together in new ways after open coding, by making connections between categories” (Strauss & Corbin, 1990, p. 96). This second step entails to link codes “to contexts, to consequences, to patterns of interaction, and to causes” (Bryman, 2012, p.
578). It is aimed to identify relationships which can be drawn between the several open codes. By doing so, axial coding can be described as the second level working towards the process of theoretical development. In this step, such relationships, however, are not only identified but furthermore rear-ranged in order to get a hierarchical order between the different codes, so that sub-categories emerge.
Those, in turn, need to be verified against the conducted data (Saunders et al., 2016).
Again, an exemplarily overview of the axial coding is illustrated in the table below (see table x).
AXIAL CODE OPEN CODE RESPECTIVE QUOTE(S)
Actual self
Boosts self-confi-dence
“Well, first off, I get the brand, the anomaly that this kid from a low-in-come suburb in the Philippines is wearing it. There’s that number one.
Because people like me are not supposed to be wearing these things or can purchase these things. So that’s cool. I call that swag.” (Int. 3)
“I think, I like of course the quality of the product material, the prod-uct itself, but also maybe the feeling that the brand adds on to you, I guess. So, I guess, the brand association kind of gives you a boost in personality or like it adds on something to your, to your sense of self.”
(Int. 5)
“I think it affects me more than I think. Because like I said, it kind of gives you a boost indirectly.” (Int. 5)
Enriches personality
“I think, by looking at things and knowing things about a brand, kind of already give you like, oh, I’m a person that knows a lot about this brand. So, I’m this guy, I think it definitely adds on to your personality as well. Like, if you follow a brand on Instagram, that’s engagement, right? And I think by following a certain brand, kind of also indirectly gives you this changed picture of yourself because I’m the girl or guy that follows this brand. So, I’m part of that a little bit. So, it’s kind of like, this is sum of the brands that I follow. So, it’s like me, this is my per-sonality and my interests.” (Int. 13)
Ideal self Helps to get closer to desired values
“I believe very strongly, that if you transfer the values and visions of the brand to yourself through products and thereby radiate some-thing that others associate with the brand. So, if someone else has a certain image of a brand and I am wearing this specific brand, then he or she would project the values onto me, even without knowing me.” (Int. 4)
“But it’s also, I think, it’s important for the values you want to deliver and the effort you want to deliver.” (Int. 4)
Table 7: Exemplarily axial coding
4.7.2.3 Third stage: Selective coding
As the final level of coding, selective coding comes into play. Strauss and Corbin define it as “the procedure of selecting the core category, systematically relating it to other categories, validating those relationships, and filling in categories that need further refinement and development” (Strauss
& Corbin, 1990, p. 116). Thus, the aim in the final coding stage is to establish principle categories, focusing on recognizing and developing relationships between the categories created in the previous axial coding process (Saunders et al., 2016).
In the case of the research at hand, this yields into three core categories – namely luxury fashion brands, customer engagement, and lastly, customer’s self-concept. An example of the integration into selective codes can be found in the table below (see table 8).
4.7.2.3 Third stage: selective coding
As the final level of coding, selective coding comes into play. Strauss and Corbin define it as “the procedure of selecting the core category, systematically relating it to other categories, validating those relationships, and filling in categories that need further refinement and development” (Strauss
& Corbin, 1990, p. 116). Thus, the aim in the final coding stage is to establish principle categories, focusing on recognizing and developing relationships between the categories created in the previous step (axial coding) (Saunders et al., 2016).
In the case of the research at hand, this yields into three core categories – namely luxury fashion brands, customer engagement, and lastly, customer’s self-concept. An example of the integration into selective codes can be found in the table below (see table x).
SELECTIVE CODE AXIAL CODE OPEN CODE RESPECTIVE QUOTE(S)
Customer’s self-concept
Actual self
Boosts self-con-fidence
“Well, first off, I get the brand, the anomaly that this kid from a low-income suburb in the Philippines is wearing it. There’s that number one. Because people like me are not supposed to be wearing these things or can purchase these things. So that’s cool.
I call that swag.” (Int. 3)
“I think, I like of course the quality of the product material, the product itself, but also maybe the feeling that the brand adds on to you, I guess. So, I guess, the brand association kind of gives you a boost in personality or like it adds on something to your, to your sense of self.” (Int. 5)
“I think it affects me more than I think. Because like I said, it kind of gives you a boost indirectly.” (Int. 5)
Enriches per-sonality
“I think, by looking at things and knowing things about a brand, kind of already give you like, oh, I’m a person that knows a lot about this brand. So, I’m this guy, I think it definitely adds on to your personality as well. Like, if you follow a brand on Instagram, that’s engagement, right? And I think by following a certain brand, kind of also indirectly gives you this changed picture of yourself because I’m the girl or guy that follows this brand. So, I’m part of that a little bit. So, it’s kind of like, this is sum of the brands that I follow. So, it’s like me, this is my personality and my interests.”
(Int. 13)
Helps to get
“I believe very strongly, that if you transfer the values and visions of the brand to yourself through products and thereby radiate something that others associate with the brand. So, if someone else has a certain image of a brand and I am wearing this
spe-5 Findings
The following section presents an overview of the findings in regard to the research question “What are the different types of customer engagement with luxury fashion brands, and how does such cus-tomer engagement influence the cuscus-tomer’s self-concept?”. The presented data is thereby generated from the conducted semi-structured in-depth interviews and is identified in the subsequent data anal-ysis applying GT.
First, brief insights on how the participants define LFBs from their customer perspectives are given.
In order to answer the research question and following the Interpretivist world view as well as the subsequent GT approach, the developed conceptual model is presented and functions as a guide to lead through the presentation of the findings. Answering the first part of the research question, the different types of CE within the luxury fashion industry are revealed. The second part of the research question is to be answered by presenting the influence of the CE with LFBs on the customer’s SC, which in turn is split into the four different dimensions of the SC by which the construct is defined in the study at hand (see chapter 2.3).
All of the assumptions presented are supported by quotations from the several participants. Thereby it becomes crucial to mention that the number of quotes assigned to the categories does not serve as information on the qualitative value of the coded statements and is not the focus of the analysis. A complete version of the respective code list as well as of the codebook including the exact frequency of the quotes can be found in the appendix (see appendix C and D).
In order to ensure that the participants understand LFBs as something that is consistent with the def-inition in the literature, the participants were also asked as part of the interviews to define LFBs from their perspective. However, since the insights gained from this are not directly related to answering the research question, this will only be briefly touched upon here. A detailed presentation of the quo-tations regarding the definition of the LFBs can therefore be found in the appendix (see appendix C).
In this regard, analyzing the data reveals distinct characteristics of LFBs, which in turn can be as-signed to eight key categories – namely, tangible product, high-end, significant identities, beyond usual, visual uniqueness, well-being, communication, timeless vs. innovative – on the basis of which the construct LFB is defined from the consumer’s perspective.