Construct Validity and Measurement Invariance of the Difficulties in Emotion Regulation Scale - Short Form (DERS-SF): Further Evidence From Community and Student Samples

Abstract

The current study aimed to evaluate the psychometric properties of the Difficulties in Emotion Regulation Scale-Short Form (DERS-SF) in Iran, including testing its measurement invariance across sexes, as well as community and student populations. Two samples were recruited: a community sample of 583 participants (58.7% female; M_age = 33.55) and a university student sample of 409 participants (67.2% female; M_age = 24.48). Besides the DERS-SF, participants completed a battery of instruments online, measuring mentalizing capacity and borderline personality features. Confirmatory factor analyses supported the tenability of the five-factor model, excluding the awareness subscale. Except for the awareness subscale, acceptable to excellent internal consistencies were found for the DERS-SF and its subscales. The awareness-excluded DERS-SF was significantly and strongly associated with relevant constructs (|r_s| = .49 to .59). This study also found evidence for configural, metric, and scalar invariance of the DERS-SF across sexes and community and student populations. Our findings extended the evidence for the validity and reliability of the DERS-SF and its awareness-excluded version by administering it in Iranian samples and supporting its cross-cultural applicability.

Disclosure statement

The authors declare that they have no conflict of interest.

Open Scholarship

This article has earned the Center for Open Science badges for Open Data through Open Practices Disclosure. The data are openly accessible at https://doi.org/10.17605/OSF.IO/UVJSZ. To obtain the author’s disclosure form, please contact the Editor.

Data availability statement

The data that support the findings of this study are permanently and openly available in the Open Science Framework at https://doi.org/10.17605/OSF.IO/UVJSZ.

Notes

1 Maximum likelihood (ML) estimation should be employed for continuous variables. However, since ML is suggested to be suitable for Likert-type scales with 5 or more response categories (e.g., Koziol, 2023; Rhemtulla et al., 2012), and caution is recommended with ΔAFI indices when using mean- and variance-adjusted weighted least square (WLSMV) for measurement invariance testing (Sass et al., 2014), we reported the ML estimation in the main text. Additionally, the CFA analyses were also conducted using WLSMV estimation (Supplemental material 1, Tables 1 and 2). No practical differences were observed.