Assessing Implicit Evaluative Attitudes Toward Violence in Male Students, Male Community Members, and Men Convicted of Violent Offences

According to theory and research, evaluative attitudes are considered a central and precipitating factor in behaviour; however, this construct appears less focussed on within the violence literature. Evaluative attitudes can be classified as explicit (assessed with self-report scales) and implicit, often assessed using response latency measures. Some studies have reported a link between implicit evaluative attitudes toward violence, violent behaviour, and risk relevant constructs, whereas others have reported no link. The aim of this dissertation was to modify existing implicit procedures to assess implicit evaluative attitudes toward violence alongside other commonly assessed violent cognitions. Three response latency measures were administered: a traditional Implicit Association Test (IAT), Personalized IAT, and Relational Responding Task. A Pre-test was conducted to determine the optimal categories and stimuli to use for the implicit measures (N = 207 online adult males). The relationships between implicit evaluative attitudes toward violence, explicit evaluative attitudes toward violence, beliefs regarding violence, and violent behaviour were assessed among three samples of adult males: undergraduate students (N = 156) and community members (N = 95; Study 1), men convicted of violent offences (N = 33; Study 2), and community members recruited online (N = 627 and 820; Study 3). Overall, implicit evaluative attitudes toward violence were more strongly related to beliefs regarding violence than explicit evaluative attitudes toward violence. More positive implicit evaluative attitudes toward violence were consistently related to greater likelihood of violence. Explicit evaluative attitudes toward violence and beliefs regarding violence were moderately to strongly associated with one another, as well as violent outcomes. Using regression analyses, implicit evaluative attitudes toward violence consistently explained additional variance in likelihood of violence over explicit/self-report violent cognitions, as well as moderated the relationships between explicit/self-report violent cognitions and some violent outcomes (e.g., likelihood in Studies 1 and 3 and violence risk in Study 2). In Study 3, results from an exploratory factor analysis (EFA) demonstrated that each measure assessed its own cognitive construct regarding violence. Most factors from the EFA were significantly and incrementally related to violent outcomes. These findings will advance understanding of violent cognitions, their assessment, and role in violent behaviour.