Information-processing bias in social phobia

Abstract

Social phobia is a persistent disorder that is unlikely to be maintained by avoidance alone. One reason for the enduring nature of social phobia may be the way individuals with the disorder process social information. It is important for those involved in social phobia to have an understanding of information-processing biases, because it has the potential to guide psychological interventions. In this review of social phobia, probability and cost estimates of social situations are examined, interpretive biases are evaluated and findings relating to memory and negative imagery are also reviewed. The clinical implications of social-phobia-related information-processing biases are discussed and possible avenues for future research are outlined.

Introduction

Why does social phobia persist for years and even decades, if left untreated? Although important, it seems unlikely that avoidance of social situations can provide a complete account of the persistence of the disorder. Most patients with social phobia have to enter at least some of their feared social situations on a regular basis. When doing so, they rarely receive unambiguous negative feedback from other people. Despite this naturalistic exposure, their fears persist. One possible explanation for such persistence is that patients with social phobia process social information in an excessively negative manner. In keeping with this, cognitive theorists (e.g., Beck et al., 1985, Clark & Wells, 1995, Hartman, 1983, Leary, 1983, Mellings & Alden, 2000, Rapee & Heimberg, 1997, Trower & Gilbert, 1989) suggest that biases in information processing may contribute to the maintenance of social phobia. An information-processing bias is evident when an individual is characterised by a particular way they process information in a given cognitive domain (e.g., memory, interpretation). In particular, it is suggested that biases in attention, interpretation, memory, and imagery may lead patients with social phobia to see social situations as much more threatening than they really are, and to motivate them to behave in ways that contribute to the maintenance of their fears. For example, enhanced threat appraisals generated by information-processing biases may motivate safety behaviours which augment symptoms and elicit less friendly behaviour from other people (see Clark & Wells, 1995, Rapee & Heimberg, 1997, for further details). The present article reviews experimental studies of information-processing biases in social phobia, delineates the main types of bias, and makes recommendations for future research.

Within an information-processing approach, there are at least six possible explanations for the fact that patients with social phobia view social interactions as threatening and these possibilities are not mutually exclusive. Patients with social phobia may (1) make excessively negative predictions about future social events; (2) interpret ongoing social events in an excessively negative manner; (3) selectively retrieve negative information about past social events; (4) have distorted negative images of their own social performance; (5) show reduced processing of social cues; and (6) within social cues, tend to focus on information that could be interpreted in a negative fashion. Possibilities 1 to 4 are reviewed in detail below. Possibilities 5 and 6 are discussed in Bögels and Mansell (2004) in another article in this issue.

It is not the intention of this review to examine particular cognitive models, but rather to review the empirical literature on information processing in social phobia. Social phobia is associated with information-processing biases that operate when patients with social phobia are anticipating a social event, during a social situation and after leaving the social situation, when reflecting upon it.

In this review, we will focus on research that investigates such biases in patients with social phobia and other research that uses analogue (high versus low social anxiety nonclinical populations) designs. It is generally assumed that individuals with social phobia have levels of social anxiety that lie at the more severe end of the social anxiety continuum. Consistent with this suggestion, in a recent review, Stopa and Clark (2001) noted that, thus far, information-processing biases evident in studies that contrast patients with social phobia and nonpatient controls, have also been evident in analogue studies (high versus low social anxiety). Of course, such similarity may not always be observed in the future. For this reason, findings that are initially reported in analogue studies need to be confirmed in patient versus nonpatient studies. Given this point, we have always made evident the type of design used in each study. The need to confirm findings from analogue design studies in patient studies raises the issue of whether analogue designs are useful at all. The main advantage is the ready availability of participants. Patients with social phobia have a low help seeking rate, so it can take a research group a long time to recruit patients for a study, especially if it involves complex factorial designs and as a consequence necessitates a large sample size. Hence, analogue research can help establish potential information-processing biases that may maintain social phobia, thus facilitating progress in the field.

Throughout this review, the criterion for statistical significance will be set at p<0.05. Hence, when interactions are not significant, but authors report follow-up analyses, these will not be discussed.