Social Anxiety and Jumping-to-Conclusions Bias: A Bayesian Lens

Abstract

Social anxiety is characterised by apprehension about social situations that involve the possibility of being evaluated. Some research has found that social anxiety could be associated with a type of reasoning bias known as the Jumping-to-Conclusions bias. This reasoning bias refers to the tendency to decide within minimal pieces of information and is commonly investigated about clinical delusion (as part of psychosis). However, the relationship between social anxiety and the Jumping-to-Conclusions bias has been contentious with a limited understanding of the potential mechanism underlying this relationship. This thesis aimed to investigate a potential mechanism underlying social anxiety and the Jumping-to-Conclusions bias using a Bayesian framework, specifically through evidence weighting via computational modelling and empirical experiments.

A Bayesian computational model was developed to understand how the Jumping-to-Conclusions bias could occur as a result of uneven evidence weighting due to one's trait level across a variety of trait variables. The thesis also introduced two new Beads Task variants that tapped into socially neutral and threatening situations. Five studies were conducted using the Bayesian model to investigate the specificity of the relationship between social anxiety and the Jumping-to-Conclusions bias and to explore the role of fears of evaluation and causal effects of state social anxiety in modifying this relationship.

It was shown that the Beads Tasks could be administered as a repeated measure under some circumstances. More importantly, higher levels of social anxiety and fears of evaluation were found to potentially promote increased belief updating when one experiences elevated levels of social anxiety and when state social anxiety is heightened. The increased belief updating occurs as a result of significantly greater weights assigned to frequently occurring information to update beliefs. The studies also showed that an in-person study may induce a more careful performance in participants. However, the effects of social anxiety and fears of evaluation on evidence weighting are insufficient to outweigh the general cautiousness in belief updating that occurs at the population level to the extent that these variables contribute to the Jumping-to-Conclusions bias. Finally, it was found that social anxiety and fears of evaluation were not significantly associated with changes in observed behavioural responses including the number of draws to decisions or subjective certainty at the point of decisions. Thus, social anxiety may not be related to the tendency to reach a conclusion using limited evidence.

Overall, the five studies contributed to the theoretical understanding of social anxiety and the Jumping-to-Conclusions bias; social anxiety is associated with differences in weighting evidence that increased quicker belief updating patterns rather than hasty decision-making. The resulting clinical implication that arises from these findings is that intervention of social anxiety may focus on the evaluation of information rather than targeting the evidence collection to reduce the Jumping-to-Conclusions bias in social anxiety. Show more