Separating Common from Unique Variance Within Emotional Distress: An Examination of Reliability and Relations to Worry.

BACKGROUND

High comorbidity rates among emotional disorders have led researchers to examine transdiagnostic factors that may contribute to shared psychopathology. Bifactor models provide a unique method for examining transdiagnostic variables by modelling the common and unique factors within measures. Previous findings suggest that the bifactor model of the Depression Anxiety and Stress Scale (DASS) may provide a method for examining transdiagnostic factors within emotional disorders.

AIMS

This study aimed to replicate the bifactor model of the DASS, a multidimensional measure of psychological distress, within a US adult sample and provide initial estimates of the reliability of the general and domain-specific factors. Furthermore, this study hypothesized that Worry, a theorized transdiagnostic variable, would show stronger relations to general emotional distress than domain-specific subscales.

METHOD

Confirmatory factor analysis was used to evaluate the bifactor model structure of the DASS in 456 US adult participants (279 females and 177 males, mean age 35.9 years) recruited online.

RESULTS

The DASS bifactor model fitted well (CFI = 0.98; RMSEA = 0.05). The General Emotional Distress factor accounted for most of the reliable variance in item scores. Domain-specific subscales accounted for modest portions of reliable variance in items after accounting for the general scale. Finally, structural equation modelling indicated that Worry was strongly predicted by the General Emotional Distress factor.

CONCLUSIONS

The DASS bifactor model is generalizable to a US community sample and General Emotional Distress, but not domain-specific factors, strongly predict the transdiagnostic variable Worry.