354 TREATMENT OF SPECIFIC ANXIETY DISORDERS
ences made when social information is first encountered), the main problem is a failure
to exhibit a positive inferential bias for hypothetical ambiguous social scenarios that was
characteristic of the nonanxious group (Amir et al., 2005; Hirsch & Mathews, 1997,
2000). Moreover, individuals with high social anxiety can be trained to make positive
or benign interpretations of ambiguous social scenarios and this trained interpretation
bias reduces levels of predicted anxiety to an anticipated social encounter but not levels
of current state anxiety (Murphy, Hirsch, Mathews, Smith, & Clark, 2007). These
findings suggest that the interpretation bias evident when social phobics first encounter
a social situation is characterized by an inability to access positive or benign inferences
whereas later, more reflective interpretations that are based on preexisting maladaptive
beliefs show the enhanced negative threat bias (D. M. Clark, 2001; see also Hirsch &
Clark, 2004; Hirsch et al., 2006).
A final prediction of the first hypothesis is that individuals with social phobia will
exhibit a recall bias for negative or threatening information associated with past social
situations and their performance. However, Coles and Heimberg (2002) concluded in
their review that there was little support for an explicit memory bias for threat infor-
mation in social phobia with only 2 of 11 studies showing the predicted effect. In most
studies socially phobic individuals did not exhibit a clear recall bias for negative social
threat words compared to nonanxious controls (e.g., Lundh & Öst, 1997; Rapee et
al., 1994; Rinck & Becker, 2005). Moreover, a threat recall bias is not apparent when
socially phobic individuals recall more complex social passages (Brendle & Wenzel,
2004; Wenzel & Holt, 2002) or positive and negative social evaluative video vignettes
(Wenzel, Finstrom, et al., 2005). In addition memory bias for threat has not been appar-
ent in response to autobiographical memory cueing to social threat words (Rapee et al.,
1994; Wenzel et al., 2002; Wenzel, Werner, Cochran, & Holt, 2004).
In summary, there is strong and consistent empirical support for Hypothesis 1
from studies of interpretation bias to mildly negative or ambiguous social informa-
tion. However, it is still unclear whether a threat interpretation bias only occurs when
socially phobic individuals reflect back on their social experiences, with absence of a
self- enhancing positive interpretation bias characterizing the more immediate inferences
that occur when individuals encounter a social situation. Also there is little support for
an explicit memory bias in social phobia, but this could be due to failure to use exter-
nally valid stimuli or to prime the relevant emotional state at the time of retrieval (Coles
& Heimberg, 2002; Mansell & Clark, 1999).
Clinician Guideline 9.7
Cognitive therapy targets the socially phobic individuals’ tendency to make exaggerated
threat inferences when they ref lect on their social experiences and their inability to access
positive inferences during exposure to social situations.Hypothesis 2
The schematic organization in social phobia consists of core beliefs of an inadequate social
self, the threatening nature of social interaction, and a negative mental representation of how
one is perceived by others in the social situation.