Development and psychometric evaluation of a scrambled sentences test specifically for worry in individuals with generalised anxiety disorder

The tendency to draw negative conclusions from ambiguous information (interpretation bias) is prevalent across emotional disorders and plays a key role in the development and maintenance of pathological worry and anxious mood. Assessing interpretation bias using valid and reliable measures is central to empirical research. A commonly used measure of interpretation bias is the scrambled sentences test (SST), originally relating to depression. Given the association between interpretation bias and worry, we aimed to develop and psychometrically evaluate a new version of the SST with items pertaining to common worry domains for use in worry and anxiety research. In Studies 1–3 (analogue samples, combined N = 288), the new worry SST showed excellent construct validity (moderate-to-strong associations with worry and anxiety-related measures), and reliability (split-half and test-retest reliability). We confirmed construct validity in Study 4 (N = 215 individuals with generalised anxiety disorder). Furthermore, we demonstrated version specificity in analogue and clinical samples: the worry SST was associated with trait worry but not trait rumination, while the original depression SST largely showed the opposite pattern. Overall, the new worry SST is a psychometrically robust measure that may be especially useful for research into cognitive processes underpinning worry and anxiety.