Psychometric Properties of the PSWQ-A in a Community Sample of Older Adults

Among older adults, GAD is as prevalent as major depression (Blazer, George, & Hughes, 1991). As a result of scale development and norming that generally incorporates younger samples, psychometrically sound anxiety and worry instruments for older cohorts are limited. The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Brokovec, 1990) is one instrument that may be useful for assessing worry in older adults, although limitations of this scale recently were highlighted that resulted in the development of a revised version that more effectively might assess worry in older adults, the Penn State Worry Questionnaire-Abbreviated (PSWQ-A; Hopko et. al., 2003). The present study addressed limitations of previous work by administering the scale to an independent sample of 115 community dwelling older adults. Analyses revealed that data fit moderately with the single factor model proposed by Hopko et al. (2003). However, internal consistency of the measure was good (α = .89), test-retest reliability of the scale at the two-week (r = .92) and six-week (r = .95) follow-up administrations was excellent, and there was some support for the convergent and disciminant validity of the PSWQ-A with contemporary measures of anxiety and depression. Implications of using the PSWQA in clinical and research settings are discussed, as are study limitations.