Psychometric Properties of the Distress Tolerance Scale in a Clinical Sample

The factor structure, reliability, and concurrent validity of the Distress Tolerance Scale were evaluated in a large outpatient sample (N = 775). Prior research demonstrates mixed findings regarding the most appropriate factor structure, finding evidence for the presence of four subfactors as well as a potential second-order (hierarchical) General Distress Tolerance factor. Competing factor structures were compared using confirmatory factor analyses. A second-order hierarchical model with correlated residuals fit the data well, though results suggested poor factor discrimination. A bifactor hierarchical model also demonstrated acceptable fit. However, all subfactors except for Regulation demonstrated small or nonsignificant loadings and/or variances. The model was respecified with all items loading onto a General Distress Tolerance factor and three items loading onto the Regulation factor, which also demonstrated acceptable fit. In support of its concurrent validity, General Distress Tolerance was more strongly associated with neuroticism and a measure of difficulties with emotion regulation than with symptoms of anxiety and depression. The present study extends the literature by demonstrating support for a hierarchical bifactor structure and the favorable psychometric properties of the Distress Tolerance Scale in a large clinical sample.