142 Chapter 6 Table 1 summarizes, for each phase, the mean frequency (i.e., percentage of days) that both challenging behaviors were self-reported in the diaries. Furthermore, we calculated the mean frequency per phase for each of the 11 staff-hypothesized risk- and protective factors (see Appendix D). To obtain insight into what makes phases (dis)similar from each other in terms of these risk- and protective factors, we compared the mean frequency of them within each phase to the 560-day mean of that factor. We considered a phase-mean salient if it was above or below 1 SD relative to that factor’s 560-day mean. For example, salient about phase 1 (day 1 to 56) was that familial tensions occurred on 18% of days, which was relatively often, given that it is > 1 SD relative to the 560-day mean of 5%. Although Table 1 shows that the 11 frequencies of staff-hypothesized risk- and protective factors differ between phases, we find no unequivocal bivariate if-then explanation (e.g., if a phase has familial tensions, then high aggression) for either of the challenging behavioral frequencies. In addition to describing average frequencies across phases, we also analyzed bivariate associations at the within-day level (contemporaneous) and across days (lag-1). That is, whether challenging behaviors and reports of staff-hypothesized risk- and protective factors co-occurred on the same day and from day-to-day. Fisher’s exact test revealed that, across the entire 560-day timeline, freedom restricting measures were more often applied on days with aggression (OR = 5.27, 95%CI [2.82, 9.78]) or self-injury (OR = 2.72, 95%CI [1.56, 4.89]). Across the 560-day period, there were no bivariate contemporaneous associations between challenging behaviors and reliving trauma, hallucinating, receiving medical care, compliments or psychological therapy, having pain, sickness, experiencing negative affect or familial tensions, or positive interactions. On days after an implemented freedom restricting measure, our participant was more likely to engage in aggressive (OR = 4.80, 95%CI [2.58, 8.86]) and/or self-injurious behavior (OR = 1.97, 95%CI [1.67, 3.39]). On days after a psychological therapy session (dialectical behavior therapy or psychomotor therapy) she was less likely to engage in self-injurious behavior (OR = 0.36, 95%CI [0.15, 0.79]). To explore these associations within phases (and possible differences between phases), we repeated the same Fisher’s tests per phase, on both the contemporaneously and lagged timescale (484 tests; 11 themes × 2 behaviors × 11 phases × 2 timescales). The only significant associations that hold within certain phases evolve around freedom restricting measures, indicating that these measures were more likely to occur on the same day as aggression in phase 5, before days with self-injury in phase 7 and before days with aggression in phase 11. All other
RkJQdWJsaXNoZXIy MTk4NDMw