Hester Paanakker

3.7.3 Data analysis The data analysis consisted of a systematic content analysis through software-supported (MAXQDA) coding: a process of attaching distinct labels to data segments to organize, classify, and conceptualize the interview material (Friese, 2012; Miles & Huberman, 1994). In the belief that “a properly developed code is more than just a descriptive label” (Friese, 2012, p. 94), the coding system was developed largely inductively, using two-stage coding to build categories from the bottom up (see (Bazeley, 2007; Friese, 2012; Kuş Saillard, 2011). During the first stage, open coding was applied to the data to explore and create subcategories of qualities that provide “a good description of heterogeneity and variance in the data material” (Friese, 2012, p. 113). This includes uniting data segments with similar content into mutually-exclusive codes to create a methodological hierarchical coding system that reflects the data in all its facets (Friese, 2012, pp. 130-131). The next step was to find common denominators by renaming, modifying and integrating sub labels into larger overarching coding categories (Friese, 2012, pp. 130-131). “Going back and forth between data and codes” (Weiss, 1994, p. 156), this validated version was applied to the data set at large and allowed for grasping the subtleties of perceptions of craftsmanship, for comparing these across respondents, and for comparing ideal conceptions with conceptions of organizational facilitation. For the sake of providing a manageable overview, the analysis presented includes only qualities that were mentioned by at least five respondents. Concretely, this means that qualities as mentioned by respondents were inductively aggregated and classified into five main categories of values and one category of practical impediments that exhaustively capture and include the (common characteristics of) the qualities of craftsmanship mentioned by respondents: humanity, security, reintegration, efficiency, task effectiveness and task negativity. The nature of their content will be detailed in the results section that follows. 3.8 Findings 3.8.1 Prison Officers’ Ideal Conceptions of Craftsmanship Table 3.2 lists the key qualities that respondents associate with their ideal conception of craftsmanship in prison work for both cases collectively, because, interestingly, cross-case 77 Mismatch Between Ideals and Institutional Facilitation