Method of analysing

49

highest education, the other half has master degree. These criteria were set in order to make sure that no gender is under or over represented, and with the chosen age and educational background it can be assumed that all participants were – to a certain degree – exposed information about the importance of sustainability and have some sort of understanding of the topic. Moreover, we were aiming to interview individuals who are already engaged in some kind of sustainable consumer behaviour, including the purchase of sustainable cleaning products, so we can investigate what drives them towards this behaviour and what blocks them from being even more engaged. All participants have different levels of engagement when it comes to sustainability.

The interviews on average lasted between 45 minutes and 1,5 hours, depending on how engaged the consumer is, how many barriers they experience when purchasing sustainable home cleaning products, and how important they perceive the topic. All interviews followed the interview guide that was created beforehand, and all interviews were conducted in English, since the educational level of the participants allowed us to do this without language difficulties. Since during our research, COVID-19 safety regulations and precautions had to be taken into consideration, our decision was to conduct the interviews online in order to minimize the risk of infection, and to make sure that our participants felt safe during their contribution to our research. In the beginning of the interviews, the participants were familiarised with the topic of sustainability in the home care industry in order to avoid any misunderstanding that could potentially lead to errors in the data. During the research, some cases the participants were shown different products or images as part of the interview. In order to ensure the anonymity of the participants, we will be referring to the participants of this study without names. The interviewees will be addressed as Participant 1, 2, 3, 4, 5 and Participant 6.

50

available for the researchers when it comes to quantitative data, that helps them understand the collected information. Majority of researches analyse the data by finding different themes while reading through, and then find connections, meanings and differences when analysing (Veal, 2011). Therefore, the most well-known and commonly used method of analysis (Guest et al., 2012), thematic analysis is applied, which is a method to identify, define and interpret different patterns of meaning within qualitative data (Braun & Clarke, 2006).

The purpose of this analytical tool is to explore and make sense of the interviews, thus the phenomenon can be explored together with norms and behaviours. Thematic analysis can be an effective tool when the researcher aims to examine a research question that focuses on consumers’ behaviour, experiences and assumptions (Braun & Clarke, 2006). Furthermore, this type of analytical method is deemed to be the right choice to explore a certain phenomena in terms of different social behaviours, norms and meanings in different social contexts, and should not be considered as a single technique but a collection of different analysis methods (Guest et al., 2012). Another great advantage of this analytical method is the ability to examine the phenomena as it is in a larger picture rather than focusing on words and sentences separately in a flexible way (Guest et al., 2012), however, this flexibility usually requires a solid theoretical background that helps guide the researcher to extract the meaning from the data they look for. On the other hand, the researchers are encouraged to be stay open outside of the theoretical framework too, as focusing solely on viewing the data through the chosen framework could stop the researchers from spotting connections outside the score of the theoretical framework (Braun & Clarke, 2006).

Using thematical analysis carries another big benefit, that it enables the research question to guide the analysis and lead the main and relevant themes, thus the research question does not need to be fixed and can be altered during the research process. For this reason, thematic analysis was deemed relevant for this study, and created room for the research question to be improved and developed as the research moved forward. As more and more data were collected, new themes and topics emerged as they were mentioned by the subjects, and they

51

could be included in the findings and implications. Within thematic analysis, there are several approaches and procedures that can be effective in analysing different type of data.

This study implemented an approach that is a mix of two different thematical analysis methods. One of these is template analysis, which is one type of thematic analysis. This type of analysis is very popular as it is quite flexible and can be adopted to most type of qualitative research (King & Brooks, 2017). When template analysis is applied, pre-existing themes are being used that are based on the theoretical framework of the research, but the same time this method enables to be open to new themes as well that can emerge from the data that was collected. Therefore, this method works well with researches that follow abductive research approach and also utilises the benefits that semi-structured interviews as a data collection method can provide. The template analysis most commonly is being conducted in three steps;

transcription, coding, and interpretation (King & Brooks, 2017). The first step is to transcribe the interviews, this can be done during, or after the interview took place. Second step suggest a type of coding, where the gathered data is organized into themes – these are both themes that were set during the setup of the theoretical framework, and themes that emerged during data collection. This phase requires the researchers to get familiar with the collected data by reading and understanding the interviews many times. Finally, after the coding took place and the relevant themes and sub-themes were determined, the last step is the interpretation of the data.

Similar, but slightly different approach is described by Braun and Clarke (2006), who recommend a six-step analytical approach that help identify and analyse the collected qualitative data. This approach is also quite flexible, since the different steps can be conducted in different orders, so the analysts have the option to decide and create the flow how it bests fits the given research.

As a first step, in line with the suggested approach by King and Brooks (2017), the analysts need to dive into the data and learn it as much as possible in order to help with the further analysis (Braun & Clarke, 2006). All interviews were recorded and then later transcribed, thus both researchers had the opportunity to listen to all the interviews and gain as much knowledge as possible. The researchers continuously re-read the transcriptions of the interviews and had

52

discussions about the data already collected in order to further understand what aspects to focus on during the next interviews. Every interview was conducted at least a with a week break after the previous one, therefore there was sufficient time to familiarize with the last data collected for both researchers and have a discussion of it.

The second step recommends the creation of codes (Braun & Clarke, 2006), while the data is re-read. This phase is again in line with the suggested steps by King and Brooks (2017).

Thirdly, all the codes that were created in step 2 should be revised and grouped together in order to create the themes deemed most relevant (Braun & Clarke, 2006). In order to form the most relevant themes and clusters the researchers sat together and used their knowledge and insights gained while conducting the interviews and researching the topics.

As the fourth step, all the created themes need to be revised in the light of the theoretical framework and research agenda, and the researchers need to decide if the themes are deemed relevant or not (Braun & Clarke, 2006). During this step, the researchers have the possibility to make small changes to the created themes or even add more if they believe it is necessary to be compliant with the research question and theoretical framework.

During the next step, the researchers take the final themes that were considered most relevant, and properly and fully define them, clear the borders of them and name them (Braun & Clarke, 2006). Finally, in the last phase of thematical analysis, the data needs to be interpreted by the researchers in light of the themes, and the findings can be extracted (Braun & Clarke, 2006).

In this thesis, we followed the suggested steps and phases of both Braun and Clarke (2006) and King and Brooks (2017) while analysing the data we collected in order to ensure a trustworthy method of data analysis and minimize the chance of errors.