Abnormal and Behavioural Psychology

Construct Validity of the Difficulties in Emotion Regulation Scale: Further Evidence Using Confirmatory Factor Analytic Approach

Abstract

Leonardo Adrián Medrano and Mario Trógolo

Background:
The Difficulties in Emotion Regulation Scale (DERS) represents one of the most popular, comprehensive and well-established measure of emotion regulation, being widely used in clinical and nonclinical settings. Despite it widespread use, there is no agreement about the factorial structure and majority of prior research has been focused in replicate the original DERS model, while studies examining competing models are lacking. Thus, further examination is needed in order to ascertain the more appropriate factor structure. The present study sought to examine the latent factor structure of the DERS by testing several alternative models using a confirmatory factor analytic (CFA) strategy.
Methods:
Six-hundred and nine adults (367 men) from general population aged from 18 to 51 years old completed the DERS.
Results:
Findings did not support the original six-factor correlated model. Rather, CFA studies showed that a four-factor correlated model including lack of emotional clarity, no acceptance of emotional responses, difficulties engaging in goal-directed behavior, and impulse control difficulties had better fit to the data. Drawing on empirical research from affective neuroscience, a two correlated second-order CFA model accounting for the four factors was examined. The hypothesized two factors were represented by difficulties in emotion processing and difficulties in the regulation of emotional response. Results showed acceptable and similar goodness-of-fit indices compared to the four-factor correlated model, while test for model comparison revealed non-significant differences between fit of the two models.
Conclusion:
Results indicate that both a four-factor correlated model and a two higher-order correlated factor model are adequate to explain the latent factor structure of DERS. However, the last model is recommended since it provides researchers with a more parsimonious, neurobiological-based conceptualization and assessment of emotion regulation. Limitations and directions for future investigation are also addressed.