134 Chapter 6 the different attractor states. Thirdly, we evaluated transition between attractor states in terms of (in)stability and extraordinary events (cf. Figure 1). 2.4 Analytical strategy 2.4.1 Summarizing daily life The first step was to qualitatively analyze the anonymized daily records in accordance with the phased approach of thematic analysis (Braun & Clarke, 2006). This thematic analysis was conducted by the first author together with four Master’s students in Pedagogical Sciences, all under the supervision of a researcher with ample experience in qualitative methods. A thematic analysis is an inductive method whereby the coders collaboratively construct themes and patterns from the text in an iterative process that contains six phases: data familiarization, generating initial codes, searching for themes, reviewing themes, defining themes, and producing the research report. In each of these phases, the coders frequently came together to discuss and interpret the records. All five coders first familiarized themselves with the data by reading the whole daily records text file, which consisted of >300,000 words. Together the coders then practiced the initial coding. The text file was then divided into five roughly equally large chunks of daily records text. Each coder then generated initial codes on his/her own text. The coding was done using MAXQDA (2022). During the initial code generating phase, coders came together thrice to compare each other’s initial coding wording and interpretation of the text. These iterative consensus-building sessions lead to the construction of a preliminary overview of candidate subthemes and themes (i.e., codes that were interpreted as reflecting the same higher-order construct). During this collaborative, inductive process, the wording and structure of these (sub)themes were refined into one thematic overview that contained a theme- and subtheme-structure that captured themes based on the whole dataset. This procedure fosters a shared understanding among all coders, resulting in a consensus over the overarching thematic structure (thematic map). From this jointly construed thematic map, every coder then coded the records once more from scratch. That finally resulted in a MAXQDA file with fragments of coded text on a specific day. These qualitative data were then quantified to a dataset containing only binary variables with a (sub)theme coding present (1) or not present (0) per day. The researcher then met with the participant’s clinician, who has known her for over a decade, to discuss the thematic overview and underlying codes. The goal of this meeting with the clinician was to (1) ascertain the appropriateness of challenging behaviors as the most indicatory variables to summarize the system’s overall state, and (2) identify the most relevant (sub) themes for explaining the frequency of challenging behavioral incidents at any given period.
RkJQdWJsaXNoZXIy MTk4NDMw