Empirical Status of the Cognitive Model 95
belief that it helps in preparation for anticipated future negative outcomes (Borkovec
& Roemer, 1995). Wells (1994b, 1997) has argued persuasively that positive beliefs
about worry’s effectiveness in threat reduction contribute to its persistence. However,
the effectiveness of worry is immediately undermined by the fact that most of the things
that people worry about never happen (Borkovec et al., 2004). Under these conditions a
powerful negative reinforcement schedule is set in place in which positive beliefs about
the effectiveness of worry for avoiding or preventing bad events become strengthened by
the nonoccurrence of adverse events. So we worry not to gain any particular advantage
but instead to prevent or avoid anticipated adversity.
Even though worry may be a superfluous cognitive activity, its negative effect is fur-
ther compounded by evidence that its very occurrence thwarts effective problem solving.
Measures of worry are negatively correlated with certain aspects of social problem-
solving measures in both clinical and nonclinical samples (Dugas, Letarte, Rhéaume,
Freeston, & Ladouceur, 1995; Dugas, Merchand, & Ladouceur, 2005). Chronic worry
is unrelated to social problem- solving ability but more directly associated with lower
problem- solving confidence, less perceived control, and reduced motivation to engage
in problem solving (Davey, 1994; Davey, Hampton, Farrell, & Davidson, 1992; Dugas
et al., 1995). In sum, this research suggests that although pathological worry may not
be characterized by social problem- solving deficits, it probably interferes in the per-
son’s ability to implement effective solutions (Davey, 1994). In contrast, worry phenom-
ena in nonclinical populations may be associated with more effective implementation
of problem- solving responses (Davey et al., 1992; Langlois, Freeston, & Ladouceur,
2000b).
EXCESSIvE WoRRY AND THE THREAT INTERPRETATIoN bIAS
A final negative consequence of worry is that it causes one to reappraise a fear stimulus
in a more threatening manner. In a study of self- reported worriers and nonworriers
in elementary school-aged children, Suarez and Bell-Dolan (2001) found that worriers
generated more threatening interpretations to hypothetical ambiguous and threatening
situations than children not prone to worry. Constans (2001) also found that worry-
proneness 6 weeks before an exam was associated with increased estimated risk of fail-
ing the exam. These findings, then, are consistent with our proposition that worry will
contribute to a reappraisal of threat as a more severe and probable occurrence.
Negative Impact of Safety Seeking
Even though various aspects of safety- seeking have been discussed previously, it can also
be viewed as a maladaptive compensatory coping strategy. More extensive reliance on
safety- seeking behavior has been linked to the persistence of anxiety and threat- related
beliefs (see section on Hypothesis 2). Furthermore, there is some evidence of a weaker
automatic processing of safety information and a later attentional avoidance of threat.
If more direct experimentation upholds the notion that automatic processing of safety
information is less efficient in high anxiety states, then this could help explain why the
anxious person has to expend more elaborative resources in the pursuit of safety.
Anxious individuals are more likely to utilize safety- seeking behaviors as a means
of coping with anxiety than nonanxious individuals (see section on Hypothesis 2). In