Martijn van Teffelen

Chapter 7 160 were randomly allocated to one session CR, I-CR, or active control (AC) interventions. At baseline, post-intervention and at one-week follow-up changes in hostile beliefs, aggressive inclinations, state anger and hostility traits were measured. Moreover, at one-week follow-up participants were exposed to a provocation. We expected that I-CR would be efficacious in reducing hostile beliefs, aggressive inclinations, state anger and hostility traits compared to CR and AC. Findings showed that hostile beliefs, aggressive inclinations, and state anger were more strongly reduced by I-CR and CR compared to AC. Additionally, I-CR was sustainably more efficacious in reducing hostile beliefs and aggressive inclinations compared to AC. All conditions reduced hostility traits over time. These findings suggest that integrating the use of mental imagery in standard CR for hostility results in more elaborate processing, likely because mental imagery has a stronger cognitive impact. Implementation of computerized CBM-I Another way of advancing treatment options for hostility is by developing and implementing novel interventions. This is important because a significant number of patients do not benefit from treatment as usual. One intervention that holds promise to reduce hostility is CBM-I. Chapter 6 describes the development of a novel CBM-I intervention and its` treatment effect in two experiments. The first experiment tested the feasibility of the CBM-I intervention compared to an active control (AC) condition in a small mixed community-clinical male sample. The second experiment described a randomized controlled trial, comparing CBM-I to AC in a large sample of people with clinical levels of hostility. It was expected that CBM-I would result in a stronger increase in benign bias and stronger reductions of hostile bias and hostility symptoms and traits compared to AC. The second experiment also explored whether CBM-I was related to beneficial carry-over effects in case people engaged in psychotherapy after CBM-I. Findings confirmed that CBM-I efficaciously increased benign bias in both trials. Moreover, the second experiment demonstrated that CBM-I reduced hostile bias, behavioral aggression compared to AC. However, CBM-I did not reduce affective aspects of hostility compared to AC. Following the evidence presented in chapters 4 and 6 , all interventions (i.e., I-CR, CR and CBM-I) effectively reduced aggressive behavior by targeting hostile interpretations. This evidence is in line with contemporary working models of hostility (Allen et al., 2018; Crick & Dodge, 1994; Wilkowski & Robinson, 2008) that predict reductions in aggressive behavior by establishing reappraisal strategies. Our findings indicate that these reappraisal strategies can be installed in different ways. On the one hand, (I-)CR attempts to install a reappraisal strategy through the elaborate personalized evaluation of provoking situations (top-down). On the other hand, CBM-I attempts to install a reappraisal strategy by repeatedly reinforcing benign interpretations in many random emotionally ambiguous scenarios (bottom-up). Perhaps, combining both top-down and bottom-up reappraisal strategies can be complementary in the treatment of hostility.