Recent studies have enhanced the utility of the PSWQ by testing for optimal cut-scores to screen for GAD "caseness." Several studies have applied receiver operating characteristic analysis to determine PSWQ scores that optimize sensitivity (likelihood of identifying "true positives" from all persons with positive diagnosis according to the questionnaire) and specificity (likelihood of identifying "true negatives" from all persons without diagnosis on the measure) vis-a-vis clinician diagnosis or questionnaire-selected analogue GAD. Behar, Alcaine, Zuellig and Borkovec (2003) found that a PSWQ score of 45 optimized sensitivity and specificity in discriminating treatment-seeking GAD clients from non-anxious controls; in contrast, in a large student sample, a score of 62 performed best. This latter, higher, score was required to differentiate analogue GAD individuals (categorized via GAD-Q-IV) from those with self-reported diagnoses of social anxiety disorder, moderate depression, and/or PTSD. This highlights the fact that the ability of the PSwQ to identify GAD depends on the sample composition in relation to other affective symptomatology.
Similarly, Fresco et al. (2003) found that optimal PSWQ scores for identifying clinician-diagnosed GAD caseness varied depending on the goal (i.e., sensitivity, specificity, or both) and the target group from which GAD was being differentiated. For discrimination of "pure" GAD from "pure" social anxiety disorder, scores that optimized sensitivity (57), specificity (69), and both (65) were reported. In comparison, discrimination of primary or secondary GAD from pure social phobia yielded scores of 57,68, and 65, respectively. Chelminski and Zimmerman (2003) reported convergent results in regard to a score that provides the best balance of sensitivity and specificity in differentiating GAD patients from non-GAD patients, arriving at a score of 64 in a large outpatient sample. It is noteworthy that several of these optimized cut-scores fall in a comparable range with aforementioned means for GAD patients.
In sum, the PSWQ demonstrates the ability to discriminate GAD, but this capacity depends upon the context and purpose for utilizing the questionnaire; clinicians and researchers should choose cut-scores based upon consideration of their goals. If the aim is to avoid failing to detect individuals with GAD symptoms, lower scores may be used; to avoid false positives, higher scores are recommended. Furthermore, higher scores are required to discriminate GAD from other conditions prone to heightened worry, whereas lower scores suffice when the comparison group is a non-anxious sample.
Was this article helpful?