Empirical Status of the Cognitive Model 77
Wells & Mathews, 1994). Metaworry involves a subjective negative appraisal of the
significance, increased incidence, and perceived difficulties associated with the uncon-
trollability of worry (Wells & Mathews, 1994). Evidence that GAD is associated with
heightened metaworry would support Hypothesis 5 and indicate that in generalized
anxiety a negative interpretation of the act of worrying (e.g., “If I don’t stop worrying,
I’ll end up an emotional wreck”) contributes to an intensification and persistence of the
anxious state. In fact, various studies have shown that GAD patients were distinguished
from patients with other anxiety disorders (especially social phobia) by heightened
scores on metaworry (Wells & Carter, 2001) and there is a strong relationship between
metaworry and increased tendency to experience pathological worry (Wells & Carter,
1999; Wells & Papageorgiou, 1998a; see also Rassin, Merchelback, Muris, & Spaan,
1999). An early study by Ingram (1990) found that generalized anxiety and depression
were characterized by a heightened focus on one’s thoughts, sensations, and feelings
as indicated by Fenigstein, Scheier, and Buss’s (1975) Self- Consciousness Scale (SCS).
These studies are consistent with Hypothesis 5, indicating that an increased focus on the
negative characteristics of worry will exacerbate the general anxiety state.
In social phobia negative interpretation of anxious symptoms in social situations
because of a concern that anxiety will be perceived negatively by others is a central
feature of the disorder (see D. M. Clark & Wells, 1995; Wells & Clark, 1997). Various
studies have found that social phobia is characterized by negative appraisal of anxiety-
related interoceptive cues that leads to erroneous inferences about how one appears to
others and subsequently to heightened subjective anxiety (for review, see D. M. Clark,
1999; Bögels & Mansell, 2004). Elevated self- focused attention has been found in social
anxiety (e.g., Daly, Vangelisti, & Lawrence, 1989; Hackman, Surawy, & Clark, 1998;
Mellings & Alden, 2000). Moreover, a specific focus on anxious symptoms (e.g., blush-
ing) intensifies anxiety in high social anxiety but not in low social anxiety (Bögels &
Lamers, 2002; see Bögels, Rijsemus, & De Jong, 2002, for contrary findings).
Experimental research has also supported the cognitive model. Mansell and D. M.
Clark (1999) found a significant association in high but not low social anxiety between
perception of bodily sensations and ratings of how anxious individuals thought they
appeared to others. Mauss, Wilhelm, and Gross (2004) compared high and low socially
anxious students before, during, and after a 3-minute impromptu speech and found that
the high social anxiety group perceived a greater level of physiological arousal, felt more
anxious, and exhibited more anxious behavior than the low anxious group even though
there were no significant group differences in actual physiological activation. Moreover,
self- reported anxiety correlated with perceived but not actual physiological activation
for the total sample. These findings are consistent with Hypothesis 5. Social phobia is
characterized by a heightened focus on anxious symptoms that clearly intensifies the
anxious state.
In cognitive accounts of OCD the central problem is the faulty appraisal of
unwanted intrusive thoughts, images, or impulses of dirt, contamination, doubt, sex,
causing injury to others, and the like (D. A. Clark, 2004; Salkovskis, 1989, 1999; Rach-
man, 1997, 1998, 2003). Thus obsessional thinking develops when an unwanted intru-
sive thought, image, or impulse is misinterpreted as representing a significant potential
threat to one’s self or others and the person perceives a heightened sense of personal
responsibility to prevent this anticipated threat. Rachman (1998) suggested that “emo-