Data analysis and validation Clause Samples

Data analysis and validation. This section will present a more detailed explanation of the steps taken when triangulating the gathered data in the various databases. Personal memos and notes were written to track and review observations throughout the data analysis process. According to ▇▇▇▇▇▇▇ (2011, as cited in Yin 2018, p. 168), "memos can be attached to the computer codes and contain hints, clues, and suggestions about how to interpret some part of your data". The data analysis was conducted by triangulating the collected data using computer-assisted tools and performing manual analytical steps. The software used for the qualitative data analysis was ▇▇▇▇.▇▇. ▇▇▇▇.▇▇ allowed for clustering, coloring and flexibility with the drag-and- drop functions. The transcripts from the interviews were then systematically read, coded, and analysed through a series of thematic analysis (TA) steps, as suggested by scholar ▇▇▇▇▇▇ (2015). Step 1: The audio-recorded material was transcribed into word files using Word's transcribing functions and then manually revised and controlled to ensure that all words used in the interview were documented. Step 2, 3, 4, 4a: A deductive approach was most suitable for the data and purpose of this research, as suggested by ▇▇▇▇▇▇ (2015) since this approach allows for observing and identifying new patterns and structuring the answers in a GAP analysis structure. Therefore, the coding in these steps was developed by their direct relation to the research question, answering the question of "How". All the data in the transcribed text file was sectioned in a way that allowed for easy identification of which answer belonged to which question in the interview protocol. Then, sections of the Swedish transcripts were translated into English and written directly into the ▇▇▇▇.▇▇ file. In ▇▇▇▇.▇▇, all labelling of the codes was constructed in relevance to answering the research question. For example, words describing a function, process or how something works/and does not work were labelled as answering any of the two research questions. According to ▇▇▇ (2018), this procedure assigns various codes to the data, representing a concept or abstraction of potential interest (p.169). The data analysis and coding happened entirely in ▇▇▇▇.▇▇, where its visualisation functions were used to make the different connections to the labels and codes (see figure 5) visible.