The effect of situational structure on the social performance of socially anxious and non-anxious participants

Abstract

Twenty-six socially anxious and 24 low-anxious female undergraduate students were observed in brief unstructured and structured hetero-social interactions with a confederate. The unstructured interaction was a naturalistic interaction in which participants were observed surreptitiously. The structured interaction was a role-play in which participants were instructed to try and get to know as much as possible about their partner. Videotapes of the interactions were subsequently rated on subjective and objective measures of social skill. The results showed that high socially anxious females performed somewhat worse than low socially anxious females in both situations. However, this difference was far larger in the unstructured social situation and was relatively small in the structured social situation. It appears that socially anxious females do perform more poorly in social interactions than do low-anxious females, but a large component of this poor performance may be a result of avoidance rather than a lack of ability.

Introduction

In an early theory of social anxiety, Curran (1977) suggested that “the source of the anxiety experiences in heterosexual-social interaction is partially reactive and due to an inadequate or inappropriate behavioral repertoire. An individual may never have learned the appropriate behavior (at least to sufficient strength for smooth application) or may have learned inappropriate behaviors. Consequently, given this inadequate repertoire the individual does not handle the demands of the situation appropriately and experiences an aversive situation that elicits anxiety” (Curran, 1977, pp. 141–143).

More recent models of social anxiety have focussed on the interpretational or cognitive aspects of the condition (e.g., Clark & Wells, 1995; Rapee & Heimberg, 1997; Schlenker & Leary, 1982). According to these theories, the crucial determinant of anxiety in socially anxious individuals is an appraisal that their performance will not match a standard expected by their audience. While it is possible that this discrepancy will be produced by accurate appraisal of poor performance, it is generally assumed that it is the appraisal rather than the performance that is distorted. This perspective is supported by the numerous studies showing that socially anxious individuals perceive their own performance as considerably worse than it is perceived by an audience (e.g., Alden & Wallace, 1995; Rapee & Lim, 1992; Stopa & Clark, 1993).

To some extent, then, the question of whether socially anxious individuals lack social skills is moot. However, this is not entirely the case, since the social skills hypothesis does hold implications for treatment. If people who are socially anxious commonly lack the ability or skills to interact appropriately with others, then lasting treatment must include detailed training in these skills. Indeed, several studies have shown that treatment for social anxiety that includes a component of social skills training is effective (Turner, Beidel, Cooley, Woody, & Messer, 1994; Wlazlo, Schroeder-Hartwig, Hand, Kaiser, & Munchau, 1990). However, other studies have also demonstrated effective treatment of social anxiety without a clear skills training component (Heimberg et al., 1990; Wlazlo et al., 1990). Even attempts to match so-called “behavioral responders” to treatment focusing on social skills have failed to provide marked increments in treatment outcome (Mersch, Emmelkamp, Bögels, & van der Sleen, 1989). This would suggest that social skills’ training is not a necessary component of treatment. Nevertheless, the point remains that if people with social phobia are shown to lack basic ability in social interactions, then the most lasting and effective treatment must include some attempt to improve social skills. Thus, the question of whether socially anxious individuals lack skills is a relevant one.

A direct approach to assess social skills deficits in people who are socially anxious has been attempted in many studies over the years. Unfortunately, these studies have yielded conflicting results making conclusions impossible (Rapee, 1995). Every possible combination has been demonstrated. A few studies have shown that socially anxious individuals perform worse, both globally, and on specific behavioral measures, than do participants low in social anxiety (Twentyman & McFall, 1975). More commonly, several studies have failed to find strong differences between socially anxious and low-anxious individuals on individual behaviors, but have shown a difference only on broad, global assessment of performance (Beidel, Turner, & Dancu, 1985; Borkovec, Stone, O’Brien, & Kaloupek, 1974). Finally, some studies have failed to find significant differences between socially anxious and low-anxious individuals on either specific behaviors or on global measures of performance (Clark & Arkowitz, 1975; Rapee & Lim, 1992; Rehm & Marston, 1968).

Interpretation of these results is difficult. One possibility is that socially anxious individuals do lack the ability to perform adequately in social situations but that some studies have lacked sufficiently detailed measures or the power to detect these deficits. However, even demonstration of a difference between groups on social performance does not necessarily indicate a lack of skills and abilities. Rather, it is possible that socially anxious individuals may show poor social performance as a result (rather than a cause) of heightened anxiety. If this is the case, it would suggest that socially anxious individuals may have the requisite skills but cannot use them adequately under certain circumstances due to the interfering effects of anxiety. An alternative but very similar view would suggest that they fail to show appropriate social performance as a type of avoidance behavior. For example, if an individual believes that they are unskilled (even though they may not be), they may not attempt to invoke those skills in order not to be proven incompetent. In turn, this would have the effect of reducing social interaction and appearing incompetent. Recent models of social phobia have referred to this type of function as subtle avoidance (Rapee & Heimberg, 1997) or safety behaviors (Clark & Wells, 1995).

From this perspective, one could suggest that socially anxious individuals may show different levels of social performance across different situations. Thus, one possible explanation for the widely varying findings between studies of social anxiety and social performance may be variations in the situations where the performance is based. As a specific hypothesis, it could be suggested that situations that hold unclear demands and expectations for performance will produce poor performance in socially anxious individuals due to the potential for subtle avoidance. On the other hand, situations that explicitly demand performance will show less difference since they do not allow the individual to avoid. Along these lines, Leary (1983) suggested that social situations might be represented by a continuum bound by unstructured and structured interactions. Unstructured social interactions refer to those in which social roles and rules are ambiguous and structured interactions refer to those in which social roles and rules are immediately obvious and clearly defined.

Few studies have attempted to test this hypothesis directly. In one study, Nietzel and Bernstein (1976) exposed 40 unassertive undergraduates to role-play of situations requiring assertive responses. Half the participants were given “low demand” instructions in which they were told to respond as they normally would and half were given “high demand” instructions in which they were told to do the best they possibly could. Results showed that participants were able to improve their performance on demand, suggesting a performance rather than skills deficit. The lack of an assertive comparison group limits interpretation.

Pilkonis (1977) compared high and low socially anxious individuals on their performance across two different tasks. One task involved delivering an impromptu speech to an audience while the other task involved an interaction between the participant and a confederate in a waiting room. Because the waiting room situation did not include any instructions to act, this “naturalistic” situation was considered to be unstructured. On the other hand, the speech included a clear demand to perform and so was considered to be structured. Socially anxious participants performed more poorly than low-anxious participants in the waiting room task, but there was no significant difference between the groups in the speech.

The main limitation of this study was the fact that the two tasks differed not only on their implicit demands (structure), but also on the actual requirements of the task (speech vs interaction). A more careful test of the hypothesis would utilize essentially equivalent tasks that differed only on the implicit demands. This was the purpose of the present study. High and low socially anxious participants were exposed to two situations requiring hetero-social interaction. In one situation, participants were given no explicit instructions and were, therefore, able to subtly avoid interaction if desired. In the other situation, participants were explicitly instructed to attempt interaction.