416 TREATMENT OF SPECIFIC ANXIETY DISORDERS
score of 5.7 was optimal for identifying GAD from other anxiety disorder groups and
that the GAD-Q-IV correlated positively with measures of worry. In their review Turk
and Wolanin (2006) concluded that the GAD-IV-Q is sensitive to GAD but it may over-
diagnose the disorder, especially in more urban, ethnically diverse populations. In clini-
cal practice the GAD-Q-IV is unnecessary if an ADIS-IV or SCID-IV is administered.
A copy of the GAD-Q-IV can be found in Newman et al. (2002).
Measures of Worry
Penn State Worry Questionnaire
The 16-item Penn State Worry Questionnaire (PSWQ; Meyer et al., 1990) is the most
widely used measure of worry with a cutoff score of 45 recommended to distinguish
pathological worry in a treatment- seeking population (see Chapter 5 for further discus-
sion). The PSWQ should be included in the standard assessment for GAD and should be
readministered at posttreatment given its sensitivity to treatment effects.
Worry Domains Questionnaire
The Worry Domains Questionnaire (WDQ; Tallis et al., 1992) is a 25-item measure
of worry content that assesses extent of worry in five domains: relationships, lack of
confidence, aimless future, work, and financial matters. Items are assessed on a 0 (“not
at all”) to 4 (“extremely”) scale, with a total score and subscale scores for each domain
calculated by summing across respective items. The questionnaire displays good tempo-
ral reliability exhibits convergent validity with the PSWQ and trait anxiety, and GAD
samples score substantially higher than nonclinical controls (Tallis, Davey, & Bond,
1994; Stöber, 1998). Moreover the five- factor structure of the WDQ has been replicated
(Joorman & Stöber, 1997) and the WDQ correlates significantly with peer and self-
ratings of daily worry (Stöber, 1998; Verkuil, Brosschot, & Thayer, 2007). However,
the WDQ has some limitations for clinical practice. The questionnaire reflects some
aspects of constructive or adaptive worrying and so should not be construed as a “pure”
measure of pathological worry like the PSWQ (Tallis et al., 1994; Turk et al., 2004).
In addition, only certain subscales may be specific to GAD (Diefenbach et al., 2001)
and responses may be influenced by age and ethnicity (Ladouceur, Freeston, Fournier,
Dugas, & Doucet, 2002; Scott et al., 2002). A 10-item short form of the WDQ has been
published (Stöber & Joorman, 2001) and a copy of the original 25-item WDQ is avail-
able in Tallis et al. (1994). The WDQ is primarily a research instrument but can be used
clinically as a complementary tool for assessing worry content.
Cognitive Measures of Worry
Anxious Thoughts Inventory
The Anxious Thoughts Inventory (AnTi; Wells, 1994a) is a 22-item questionnaire
designed to assess both worry content and negative appraisal about worry (i.e., meta-
worry). The AnTi has three subscales: social worry, health worry and metaworry.
Although all three subscales correlate with the PSWQ, only the AnTi Meta-Worry sub-
scale shows a unique relationship with pathological worry and significantly discrimi-