Marianne Welmers

A Systemic Perspective on Alliances and their Relation to Outcome 119 CHAPTER 5 Table 1 Descriptives for CBCL scores Baseline (Mean, SD) Follow-up 1 (Mean, SD) T p Internalizing Problems 16.172 (9.385) 13.240 (8.339) 1.797 .083 Externalizing Problems 23.621 (8.809) 14.550 (8.153) 5.322 <.001 Total Problems 70.517 (21.827) 49.000 (22.123) 4.212 <.001 Note. 1 Follow-up was measured 18 months post-treatment Next, we calculated means and standard deviations of all alliances of individual family members, alliance differences, and shared sense of purpose within the family (Table 2). Therapist-, self- and observer reports all showed strongest alliances for mothers as compared to fathers, and subsequently youth, although differences were relatively small, especially between fathers and mothers. All alliances became somewhat stronger from T1 to T2, except for fathers’ observed alliances. Differences between family members’ alliances with the therapist were largest for observer reports. Interestingly, although differences between family members’ therapist- and observer-reported alliances slightly increased from T1 to T2, differences in self- reported alliances decreased. Finally, there was a small increase of shared sense of purpose from T1 to T2. We also calculated zero order and partial correlations for each family member’s individual alliance with the therapist and CBCL scores at follow-up. The results are reported in Tables 3 and Table 4. Most correlations between mothers’ and fathers’ alliances and follow up CBCL scores were small to moderate and not significant. The partial correlation between mothers’ self-reported alliance at T2 and youth externalizing behavior problems at follow up was trend-significant ( r = .277, p = .090), indicating that stronger alliances with mothers at mid-treatment were associated with more externalizing behavior problems of the child after controlling for baseline externalizing behavior problems.