Skip to content
The Kids Research Institute Australia logo
Donate

Discover . Prevent . Cure .

Properties of the DASS-21 in an Australian Community Adolescent Population

This study evaluated for a multifactor structure in the DASS-21 teenagers and the specifics of the 3 subscales for teenagers in general at different stages.

Citation:
Shaw T
, Campbell MA, Runions KC, Zubrick SR. Properties of the DASS-21 in an Australian Community Adolescent Population. Journal of clinical psychology. 2017;73(7):879-92.

Keywords:
Adolescent; Anxiety; Bifactor model; DASS-21; Depression

Abstract:
Background: Although developed for adults, the Depression Anxiety Stress Scales-Short Version (DASS-21) has been used in many research studies with adolescent samples. Evidence as to the applicability of the DASS subscale scores to represent the distinct states of depression, anxiety, and stress as experienced by adolescents is mixed, and the age at which it may be possible to differentiate these 3 states using the DASS-21 has not yet been determined. Objective: This study evaluated evidence for a multifactor structure in the DASS-21 in adolescents and the specificity of the 3 subscales for adolescents in general and at different ages. Method: Data were from a large cross-sectional survey of 2,873 school students in Grades 6–12 (aged 12–18 years) in Australia. We conducted confirmatory bifactor analyses testing a general mental health distress factor and 3 domain-specific factors for anxiety, depression, and stress for the whole sample and across gender by age groups. The internal consistency reliability of the DASS total and subscale scores was determined using omega coefficients. Results: Analyses identified that most of the variation in the items was explained by the dominance of a single, general factor and the subscales lacked specificity across all age groups. Conclusion: The DASS-21 can be reliably used to measure general distress in adolescents, but the subscales fail to discriminate between the 3 states. Our results indicate that this lack of discrimination does not reduce with increasing age. These findings caution against the use of adult theoretical models and measures within adolescent populations.