Avoidance behaviour: An investigation of overgeneralisation and ambiguity using trace procedures.

Anxiety is the most common psychiatric disorder affecting over 60 million people in Europe alone, and these numbers have increased with the ongoing pandemic. A hallmark of all anxiety disorders is avoidance behaviour, which is a diagnostic criterion for anxiety according to the DSM-V. Avoidance behaviour has been overlooked in the literature. However, a phenomenon that has received attention in relation to anxiety disorders is overgeneralisation. Overgeneralization particularly occurs in those with anxiety when there is ambiguity about potential threats. This ambiguity can occur when there is a temporal gap (i.e., trace) between predictors and dangerous outcomes. There are no assessments of generalisation with trace procedures in humans therefore highlighting that combining trace procedures with an assessment of overgeneralisation in avoidance behaviour may be of high research interest.

This project aims to investigate generalisation of avoidance performance in humans, using a novel trace procedure approach. This project will use a pre-existing avoidance task in which participants learn to discriminate between different predictive stimuli (blue and green) of an aversive outcome (loud noise). Following successful discrimination, participants will be presented with all intermediate values of the blue-green dimension and tested for generalisation. In different groups the stimulus-outcome trace will be manipulated in order to manipulate the ambiguity. It is expected that more ambiguity between the stimulus-outcome temporal relationship will lead to overgeneralisation, resulting in more avoidance and this will be mediated by participant's anxiety.

If overgeneralisation is observed then the psychological mechanisms will be investigated. One possibility is that due to the temporal gap participants do not remember all the details of the predictor resulting in overgeneralisation this phenomenon is known as forgetting of stimulus attributes. Perceptual discrimination plays a large part in generalisation and poor discrimination can lead to broader generalisation gradient. To assess this avoidance response will be measured and participants will be asked to identify the stimulus that was previously followed by the aversive outcome. This will lead to the understanding of how stimulus attributes are recalled and which elements are best identified. Finally computational modelling will be conducted to identify the different parameters that affect overgeneralisation of avoidance. Understanding the underlying mechanisms for stimulus generalisation and incorporating them into existing models will allow an assessment of the individual differences that lead to overgeneralisation.

In summary investigating overgeneralisation in avoidance could further our understanding of the underlying mechanisms which could in turn lead to treatments being adapted to attenuate relapse following treatment. Ultimately, this could result in more successful treatments for this highly prevalent disorder.