Martijn van Teffelen

Interpretation bias modification for hostility 137 6 Χ 2 = .03, p = .871 and Little’s MCAR test indicated that they were missing completely at random Χ 2 = .49, p = .975. Results are presented including participants who did not complete at least 75% of sessions while not under the influence of alcohol or drugs. The pattern of results did not change when participants who not completed at least 75% of sessions while not under influence were excluded. We present the results for bias, state hostility, general psychiatric symptoms, trait hostility, and working alliance below. Table 2 Study 2 - descriptives CBM-I ( n = 61) AC ( n = 74) Pre Post Pre Post Bias SIP hostile bias 15.90 (9.17) 11.37 (8.67) 15.51 (8.15) 14.65 (8.68) SIP benign bias 11.56 (4.42) 14.90 (4.46) 11.37 (4.07) 11.73 (4.46) WSAP hostile bias 3.43 (.81) 2.40 (1.04) 3.41 (.92) 3.15 (.88) WSAP benign bias 3.81 (.99) 4.36 (.84) 3.63 (.93) 3.67 (.72) State measures Hostile thoughts (AQ-HS) a 16.68 (8.24) 16.05 (9.01) 19.02 (8.50) 15.20 (8.48) Anger (STAXI-2S) a 29.27 (10.87) 24.32 (8.82) 28.00 (10.45) 23.59 (11.67) Aggressive behavior (FOAS) a 15.93 (3.81) 14.20 (3.74) 14.82 (3.43) 14.18 (4.02) General symptoms (K10) a c 34.95 (10.75) 37.56 (12.21) 34.78 (11.07) 40.67 (10.34) Trait hostility Overall trait hostility (PID-5H) 1.74 (.50) 1.52 (.55) 1.71 (.58) 1.57 (.68) Hostile intent (AQ-H) 23.24 (6.72) 15.66 (7.66) 23.76 (6.60) 15.51 (8.19 SR Trait aggression (FOA) a 71.73 (19.88) 53.73 (13.69) 66.41 (13.60) 51.06 (13.95) Aggressive behavior (VDT) a 10.17 (10.93) 7.00 (8.99) 11.08 (12.14) 11.22 (13.35) WAI b - 157.13 (11.10) - 146.70 (20.34) Note . a non-normally distributed. b FU measurement for people who engaged in therapy, n CBM-I = 8, n AC = 9. c Higher scores indicate less psychiatric symptoms. SR = self-reported. Bias To test the main hypothesis that CBM-I results in a larger increase in benign bias and a larger decrease in hostile bias two mixed regression models were run. The overall (i.e., effect- coded) interaction effects were significant ( p ’s < .001). Fixed (i.e., reference-coded) effects of benign and hostile bias are presented in Table 3. The effects of WSAP benign and hostile bias are shown in and Figures 2 and 3. In Table 3, time and condition variables were reference coded using the AC condition as reference. Hence, fixed effects presented in Table 3 are estimated using AC as reference category, AC baseline measurement as intercept. Findings showed that