Martijn van Teffelen

Imagery-enhanced cognitive restructuring: Efficacy 87 4 Table 3 Continued B SE t p I-CR S2 pre-stressor (M5) .61 7.31 .08 .933 I-CR S2 post-stressor (M6) 5.97 8.33 .72 .476 CR S1 baseline (M0) -7.47 2.66 -2.81** .006 CR S1 pre-intervention (M1) 6.82 6.43 1.06 .292 CR S1 image transformation (M2) 2.88 8.34 .35 .731 CR S1 post-intervention (M3) -17.90 7.15 -2.50* .014 CR S2 baseline (M4) 1.85 3.16 .58 .561 CR S2 pre-stressor (M5) -.69 7.26 -.10 .925 CR S2 post-stressor (M6) 4.60 8.28 .56 .580 Note . In all models the active control condition was chosen as reference category. a To gain additional degrees of freedomwe posed the constraint that conditions did not differ at baseline on the independent variable, hence all conditions at baseline share the same intercept; Χ 2 tests revealed that models did not significantly differ with and without this constraint ( p ’s > .260). b Intercept for the anger VAS model. S1 = session 1; S2 = session 2. 0 10 20 30 40 50 60 70 80 90 S1 baseline (M1) S1 post (M3) S2 baseline (M5) S2 post (M6) Hostile belief Time I-CR CR AC Figure 2 Estimates of hostile belief rating over time per condition Note. Error bars represent standard errors of the estimates. S1 = session 1; S2 = session 2. We only present one figure to illustrate the basic pattern of results.