Abstract

Social-cognitive theories of aggression stipulate that aggressive people have an attentional bias for aggressive and ambiguously aggressive cues. This biased social information processing is thought to occur from very basic attentional processes (encoding) through to higher order interpretative processes (representation). The present research was a detailed investigation into the relationships between aggression-related personality dimensions in young adults and attention toward images depicting general violence, intimate partner violence, and non-violent images. Participants completed measures of trait aggression, intimate partner violence perpetration and victimization, alcohol use, psychopathy, empathy, and insecure adult attachment. In a dual-picture free-viewing eye tracker paradigm, participants viewed three trial types for 2000 ms: general violence versus neutral cues; intimate partner violence versus neutral cues; and intimate partner violence versus general violence. Experiment 1 (N = 127) showed a few of the predicted relationships between the traits and attention, but Experiment 2 (N = 127) failed to replicate these findings and there was no overlap in significant results between studies. These data provide very little support for attentional biases in a healthy population toward unambiguously violent stimuli as a function of aggression-related traits.

Introduction

In an effort to understand cognitive processes underlying individual differences in aggressiveness, visual attention has become a target of study. Social information processing theory suggests that individual differences in reactive aggression are partially attributable to conscious and unconscious attentional processes (Crick & Dodge, 1994). One implication of this theory is that visual attentional processing may differentiate more aggressive people from less aggressive people. Social information processing theory posits several cognitive and behavioral stages that unfold sequentially. The first stage is the encoding process, which involves perception and attention toward social cues. During the second stage, the representation process, interpretation of the social cues takes place.

The use of different eye tracking measures that assess more automatic (encoding) versus more reflective (representation) processes may shed some light on how attention operates during each of these stages. For instance, automatic biases may be apparent in the image that captures attention (i.e., first fixation), how long before one fixates on an image (i.e., time to first fixation), and how long they look at this first attention-grabbing image (i.e., first fixation duration). Interpretation is a relatively more conscious process. In the present research, we examined the relationship between aggression-related traits and eye movements indicative of encoding and interpretation of unambiguously violent images.

General discussion

The aim of this research was to identify the extent to which individual differences related to aggression correlate with attentional measures assessed with eye tracking toward violent visual stimuli. Although numerous studies show that aggressive people are attracted to ambiguously aggressive images, across these two studies, there was no substantive support for the notion that people with aggressive traits are drawn to unambiguously violent images. There were few significant correlations between the trait measures and eye tracking outcomes. Furthermore, there was no replication across studies.

The correlations between the mental engagement items and dwell time suggested that the paradigm we used was a valid way to measure interest in the two violent image types. Indeed, participants who spent more time dwelling on violent or IPV images were able to consciously report this behavior. The individual difference measures did not correlate with self-reported mental engagement. Thus, whether by self-report or eye tracking, there was no support for people with aggressive traits showing a bias toward violent stimuli. People were engaged with the images but that the level of engagement did not depend on individual differences relevant to aggression.