Psychometric properties of the Depression Anxiety Stress Scales-21 in a nonclinical sample

Findings from several treatment and assessment investigations suggest that anxiety and depression tend to co-occur in clinical and nonclinical samples. The tripartite model was introduced to explain the relationship between these two constructs. The Depression Anxiety Stress Scales-21 (DASS-21) is one of several instruments that have been developed to explore dimensions of this model. This study examined the factor structure and other psychometric properties of the DASS-21 in a nonclinical sample and determine if the structure of the instrument represents the theoretical dimensions of the tripartite model. This instrument is composed of three subscales: Depression, Anxiety, and Stress with 7 items each. The DASS-21, together with validation measures, was administered to a sample of 252 students attending The University of Texas at San Antonio (89 men and 163 women, ages 18--47 years, mostly Caucasian and Hispanic). Internal consistency reliability estimates ranged from 0.84 to 0.90. Concurrent validity estimates were also tested. Contrary to an a priori hypothesis, significant gender differences were not found in responses to the DASS-21 items. Finally, using an item response theory modeling bifactor procedure, results were attained that provided stronger support for a general factor model rather than a three-factor model. Accordingly, the DASS-21 is conceptualized as a measure of general distress and not a multidimensional measure of depression, anxiety, and stress. Consistent with data from the current sample, it was concluded that the DASS-21 is not a representative multidimensional measure of the tripartite model. More detailed study findings, implications, and limitations are discussed.