Cognitive Processing of Attitudinal Ambivalence: Insights from a Diffusion Model Analysis

Ambivalent attitudes are characterized by positive and negative evaluations, are common in people's lives, and have implications for decision-making and evaluations. One robust finding is that response times are longer for ambivalent attitude objects than univalent ones. However, the underlying processes contributing to this effect remain unclear. Using a diffusion modeling approach, this study examines different explanations for slower responses to ambivalent attitudes. We hypothesized that information uptake is slower for ambivalent stimuli than non-ambivalent ones. Additionally, we hypothesized that trait ambivalence, a chronic state of ambivalence, correlates positively with response caution. Participants evaluated ambivalent and univalent attitude objects in a speeded binary decision task. The diffusion model parameters were estimated using a hierarchical Bayesian approach. Results revealed that response times were significantly longer for ambivalent stimuli than univalent ones. Importantly, the diffusion modeling approach revealed that information uptake was indeed slower for ambivalent attitudes objects, as shown by the lower drift rate. Furthermore, participants gathered more information for ambivalent stimuli than univalent stimuli before initiating the response as indicated by an increased boundary separation. We found no relationships between trait ambivalence and response caution. The results contribute to a better understanding of ambivalence and its cognitive processes.