Martijn van Teffelen

Chapter 7 158 Feeding the Flame: Provoking Hostility One way to study hostility is by experimentally provoking it. In light of the conceptual problem hostility currently faces (outlined above), different provocation methods may differentially impact hostility dimensions. Experimental literature offers many different ways to provoke hostility. However, different hostility induction methods were never directly compared to each other. Consequently, it was unclear which method is most effective in a given domain, limiting the generalizability of observed effects. Because of this gap in the current literature, we in chapter 3 directly compared two laboratory provocation methods that induce hostility. To reduce the confounding procedural influence, we selected two procedures with high methodological similarity: social exclusion and insult. We expected that social exclusion and insult would equally impact aggressive behavior, general negative affect, and threat perception. In addition, we added two personality concepts with potentially contrasting effects on hostility, i.e. psychopathic traits that were previously shown to reduce provoked hostility, and narcissistic traits that were previously shown to increase provoked hostility. Thus, relationships between provoked hostility, psychopathic and narcissistic personality traits were explored. Results confirmed our hypothesis that social exclusion and insult equally impact aggressive behavior, negative affect, and threat perception. Findings also revealed that under circumstances of agentic threat (i.e., insult) narcissistic traits are related to increased threat perception. In turn, threat perception was related to reduced initial aggressive responding. Results furthermore demonstrated that psychopathic traits are related to reduced negative affective change after provocation. In turn, reduced negative affective change was negatively related to provoked aggressive behavior. In other words, narcissism predisposes to feeling threatened by an insult, initially withholding aggression, but lashing out when a new threat emerges. In turn, psychopathy predisposes to experiencing less change in negative affect after provocation, leading to reduced aggressive responding. The fact that -aside from ours- no other studies directly compared the impact of hostility provocation methods is highly problematic. In order to determine if there is a true causal relationship between situational factors and hostility, scholars turn to meta-analytic studies (see e.g., Bettencourt et al., 2006; Hyatt et al., 2019). Meta-analytic studies aim to aggregate all available studies on a topic to determine a pooled effect size and are regarded as evidence with the highest quality and lowest risk of bias (Atkins et al., 2004). One of these systematic reviews across 37 studies reported comparable effect sizes for different provocation types (Bettencourt et al., 2006). However, the main methodological problem with meta-analyses in the provoked hostility field is that they have drawn conclusions on the relative effectiveness of different types of provocations without including studies that directly compare different provocations to each other, because they are absent. In a way,