A Systematic Review and Meta-Analytic Factor Analysis of the Depression Anxiety Stress Scales

[This paper is now published in Clinical Psychology: Science and Practice and can be accessed through http://dx.doi.org/10.1111/cpsp.12362.] Despite the increasingly widespread usage of the Depression Anxiety Stress Scales (DASS), its latent factor structure has yet to be reviewed. We thus carried out a systematic review of studies examining the latent factor structure of the DASS, further synthesizing findings by performing meta-analytic factor analyses on the available data from said studies. The literature search was conducted across three databases, yielding 59 articles encompassing 89 factor analyses. Reviewed results varied across factor analytic approaches (EFA vs. CFA) and DASS versions (DASS42 vs. DASS21), though the original 3-factor structure of the DASS was generally supported. However, bifactor models (consisting of a general factor and group factors) were most consistently best-fitting whenever tested, supporting an underlying unidimensional construct in DASS. This notion was corroborated by our meta-analytic factor analyses, wherein DASS21 bifactor models demonstrated the same consistent performance, though DASS42 analyses were inconclusive between a 3-factor and 1-factor solution. Our findings fit in with broader historical and contemporary developments in the field, and reinforce the potential utility of bifactor models in illustrating the overlap between depression and anxiety. Some recommendations moving forward are discussed.