جانبداری تشخیص برای چهره های انتقادی در هراس اجتماعی: تکرار و گسترش

Studies using linguistic stimuli have provided little support for explicit memory biases among individuals with social phobia (SP). However, using facial stimuli rated on their criticalness, Lundh and Öst (1996) found that individuals with SP recognized more critical than accepting faces, whereas non-anxious controls tended to show the opposite pattern. Since the publication of Lundh and Öst’s findings, additional studies using a variety of facial stimuli have produced inconsistent findings (J. Anxiety Disord. 14 (2000) 501; Behav. Res. Ther. 39 (2001) 967). Unfortunately, these inconsistencies are difficult to reconcile given great variation in methods and stimuli. Therefore, we designed a study to replicate and extend the work of Lundh and Öst (Behav. Res. Ther. 34 (1996) 787). Similar to Lundh and Öst, individuals with SP identified a significantly higher proportion of old critical faces as old than did non-anxious controls. Further, extending the work of Lundh and Öst, signal detection analyses revealed group differences on response bias according to face type. Specifically, controls showed a response bias towards indicating that accepting faces were previously seen, whereas individuals with SP did not. Finally, signal detection analyses failed to reveal group differences in the accuracy of memory.

Clinical experience and information-processing models (e.g., Williams, Watts, MacLeod, & Mathews, 1997) suggest that individuals with social phobia (SP) are characterized by biased memory for threatening information. However, a large body of research using linguistic stimuli failed to provide much support for explicit memory biases among individuals with SP. For example, in a series of four studies using a wide variety of dependent measures (recall and recognition, cued recall, recall of feedback from a hypothetical performance task, and autobiographical memory), Rapee, McCallum, Melville, Ravenscroft, and Rodney (1994) failed to find any support for explicit memory biases in SP. Similarly, Cloitre, Cancienne, Heimberg, Holt, and Liebowitz (1995) failed to find evidence for recall and recognition for social threat words among individuals with SP. In addition, two studies using self-referential encoding also failed to find differences in memory for threat and non-threat words between individuals with SP and non-anxious controls (Becker, Roth, Andrich and Margraf, 1999 and Lundh and Öst, 1997). Finally, Amir, Foa, and Coles (2000) failed to find an explicit memory bias towards socially threatening sentences in patients with SP. In contrast, Lundh and Öst (1996) found support for a recognition bias towards threat in individuals with SP. The Lundh and Öst study was unique in two important ways. First, the stimuli were pictures of faces, very externally valid stimuli for studying social anxiety. And, second, this study utilized an encoding task targeted at the core of SP, fear of negative evaluation; that is, they asked participants to rate how critical each person appeared. Individuals with SP later recognized significantly more critical faces than accepting faces, and the non-anxious controls showed a trend in the opposite direction. These findings were ground-breaking given their consistency with models of SP and that they were at odds with previous work using linguistic stimuli. However, the design of the Lundh and Öst study prohibited them from disentangling the relative contributions of memory and response bias (see Lundh & Öst, 1996: p. 792). Two additional papers examining memory for faces in SP have been published since the Lundh and Öst (1996) paper. Foa, Gilboa-Schechtman, Amir and Freshman (2000) conducted two studies examining memory for negative emotional expressions in generalized SP. In their second study, patients with generalized SP demonstrated enhanced recognition for angry and disgusted faces compared to neutral and happy faces. While these findings do show support for a bias towards negative facial expressions, their applicability to expressions reflecting negative evaluation, the core of SP, is debatable. Indeed, it is possible to imagine a situation where an observer is angry or disgusted with someone for being socially skilled or performing well. In other words, anger and disgust are not equivalent to negative evaluation. An important strength of this study is the conduct of signal detection analysis. Weaknesses of this study include lack of instructions for encoding and a small sample (n=15 patients with generalized SP). More recently, Pérez-López and Woody (2001) examined memory for facial expressions under conditions of elevated state anxiety. In a comparison recognition task, individuals with SP were not found to demonstrate an explicit memory bias towards threatening expressions. In fact, both individuals with SP and non-anxious controls demonstrated a bias towards remembering more reassuring faces. Strengths of this study include the use of a comparison recognition task to eliminate ceiling effects frequently encountered on item recognition tasks and the use of stimuli that were classified as threatening or reassuring based on an independent sample. Limitations of this study include the inability to disentangle contributions of memory and response bias and differences in the emotional intensity of the reassuring and threatening faces. Finally, given that individuals with generalized SP have been shown to divert their attention away from negative facial expressions ( Chen, Ehlers, Clark, & Mansell, 2002), the simultaneous presentation of threatening and reassuring faces in the forced recognition task may have increased the likelihood that individuals with SP would choose the reassuring faces. Finally, it is surprising that no one has sought to replicate the findings of Lundh and Öst (1996). Kazdin (1998) discusses the need for replications, stating “The importance of replication in scientific research cannot be overemphasized” (p. 409). He goes further to specify that replications are most valuable when new phenomena are being established, when they are conducted by persons other than the original investigators, and when they are done with variations in some of the methods and procedures (a partial replication of previous work, p. 412). The current study sought to replicate and extend the findings of Lundh and Öst (1996). First, using the same encoding task as Lundh and Öst, we examined whether individuals with SP would recognize more critical faces than accepting faces and whether non-anxious controls would recognize more accepting than critical faces. Next, we designed our study to allow us to conduct signal detection analyses, separating the relative contributions of memory and response bias. There are no published studies of memory for critical faces in SP that have conducted signal detection analyses. Third, we used critical and accepting faces that had been independently classified by individuals with both low and high social anxiety, thereby eliminating any possible confounding of interpretation bias, response bias, and memory bias (see Pérez-López & Woody, 2001: p. 968). We also equated the two types of faces on intensity, eliminating this as a potential confounding variable (see Pérez-López & Woody, 2001: p. 974). Finally, we obtained a larger sample than previous studies of memory for facial expressions in individuals with SP in order to increase power to detect differences.