169 COVID-19 and social relationships 6 inter-rater reliability for the coding process was good (κ = 0.90), and disagreements were resolved by consensus. Lastly, Pearson correlations were calculated and linear regression analyses were conducted to examine the effect of loneliness and social support on self-perceived changes in quality of social relationships due to the COVID-19 restrictions of forensic outpatients. More specifically, four separate forward stepwise linear regression analyses were conducted to identify possible predictors (i.e., emotional and social loneliness, positive and negative social support) of self-perceived changes in quality of social relationships (i.e., changes in general social relationships, romantic relationships, family relationships, and contact frequency). We controlled for age, gender, and daytime activities because loneliness during the COVID-19 pandemic was found to be more likely among females, within individuals under 60 years, and when daytime activities were absent (Wickens et al., 2021). Power calculations in G*Power indicated a sample size of 70 participants for multiple linear regression analyses with a single regression coefficient given an alpha of 0.05 (two-tailed) and a power of 0.80. The assumptions for multiple linear regression analyses (i.e., normality, linearity, multicollinearity, independence of residuals, homoscedasticity, and outliers) were approximately met. RESULTS In total, 70 individuals participating in the RCT were included in our analyses as they completed follow-up assessments during the COVID-19 restrictions. Only 41 participants had completed the CQ at Time 2. Demographic characteristics of the participants are presented in Table 1. Social relationships before COVID-19 To evaluate the quality of social relationships of forensic outpatients before the COVID-19 restrictions, descriptive analyses of baseline data were conducted. At baseline, the majority of these participants showed mixed to high levels of dissatisfaction on general social relationships (69%), romantic relationships (60%) and family relationships (67%). Furthermore, 53% of these participants experiences severe levels of loneliness, 40% reported moderate levels of loneliness and 7% were not lonely. We found positive social support means comparable with a normgroup consisting of a general population. However, negative social support means were significantly higher in our sample (M = 13.7, SD = 3.8), compared with a general population (M = 9.9, SD = 2.7), t(348) = 9.554, p < 0.001. This
RkJQdWJsaXNoZXIy MjY0ODMw