Elien Neimeijer

35 Discussion The aim of this study was to explore the psychometric properties of the GCI for individuals with MID-BIF who resided in a (forensic) secure treatment facility. We used conventional single level CFA to examine the factor structure of the GCI. The present study provides preliminary evidence for the construct validity and reliability of the GCI for individuals with MID-BIF. Results showed an adequate fit for a first-order and sec- ond-order model, which indicates construct validity of the GCI. Reliability coefficients for all scales were satisfactory. The support subscale loaded highest on the overall group climate scale, which indicates that support is the most important indicator of group cli- mate for individuals with MID-BIF. One item of the repression subscale (i.e., ‘Clients must ask permission for everything’ ) did not load significantly on the repression factor as a result of which it was deleted from the model to improve model fit. This finding may be related to the fact that the repression subscale had relatively lower loadings on the overall climate scale but also to the heterogeneity among the items in order to adequately capture the multifaceted nature of the construct (De Valk et al., 2016; Heynen, Van der Helm, Stams, & Korebrits, 2014; Van der Helm et al., 2011). The finding that this item is unsuitable to measure repression cannot be explained by current research. The ICC’s found in the present study indicated that a substantial portion of variance can be attrib- uted to the between-group level. In other words, the group in which each client resided accounted for a considerate proportion of the variability in perception of group climate. Our results indicate that the perception of the clients who reside in the same group is more similar to each other compared to clients from different groups. Multilevel analy- ses are recommended to explain between-group variance. There are several limitations of this study that need to be acknowledged. First, al- though the main aim of the study was to assess construct validity and reliability of the GCI for individuals with MID-BIF, client and other characteristics may be differentially related to (sub) scale scores of the GCI. Future studies should examine possible differ- ences in perceived group climate between different subgroups, addressing within-group (IQ, diagnosis, age, gender, legal status, criminal history, etc.) and between-group (secu- rity level, ward size, intensity of support etc.) variables. It cannot be ruled out that the participants did not understand some of the questions. However, because there were no drop-outs and no missing data, we believe this did not influence our results. In order to keep the level of interviewing as high as possible, monthly meetings were organised