Scrambled but valid? The scrambled sentences task as a measure of interpretation biases in psychopathology: A systematic review and meta-analysis.

The Scrambled Sentences Task (SST) is frequently used to assess interpretation biases (IBs). However, neither the range of its applications nor the quality of the empirical evidence it provides has been systematically examined. This systematic review investigates the types of samples and disorders in which the SST has been applied and evaluates its psychometric properties via a meta-analysis. The databases PubMed and EBSCOhost (including PsycINFO, PsycARTICLES, PSYNDEX, MEDLINE) were examined (last search: September 2021) and 93 studies from 91 manuscripts were included. Results showed that the SST has been applied predominantly in unselected samples or those with elevated levels of subsyndromal symptoms, with about a third of the studies employing the SST in a clinical population. While the SST was initially developed to assess depression-related IBs, it has now been extended to other disorders, in particular anxiety disorders. Results of the meta-analyses indicated good convergent validity and reliability across disorders, albeit in the context of substantial heterogeneity. Findings concerning divergent validity were mixed with high correlations across disorders between the SST and trait anxiety in particular, questioning its specificity. Future research should consider developing standardized SST versions and investigating its relationships with other measures of IB.