Meta‐analysis of the Difficulties in Emotion Regulation Scale and its short forms: A two‐part study

Abstract

AbstractThe Difficulties in Emotion Regulation Scale (DERS) is the most used self‐report questionnaire to assess deficits in emotion regulation (ER), composed of 6 dimensions and 36 items. Many studies have evaluated its factor structure, not always confirming the original results, and proposed different factor models. A possible way to try to identify the dimensionality of the DERS could be through a meta‐analysis with structural equation models (MASEM) of its factor structure. The MASEM indicated that a six‐factor model with 32 items (DERS‐32) was the most suitable to represent the dimensionality of the DERS (χ2 = 2095.96, df = 449, p < .001; root mean square error of approximation [RMSEA] = 0.024, 95% confidence interval [CI]: 0.023–0.025; comparative fit index [CFI] = 0.97; Tucker Lewis index [TLI] = 0.96; standardized root mean squared residual [SRMR] = 0.04). This result was also confirmed by a confirmatory factor analysis (χ2 = 3229.67, df = 449, p < 0.001; RMSEA = 0.075, 95% CI: 0.073–0.078; CFI = 0.94; TLI = 0.93; SRMR = 0.05) on a new sample (1092 participants; mean age: 28.28, SD = 5.82 years) recruited from the Italian population. Analyses and results from this sample are reported in the second study of this work. The DERS‐32 showed satisfactory internal consistency (i.e., ordinal α, Molenaar Sijtsma statistic, and latent class reliability coefficient) for all its dimensions and correctly categorized individuals with probable borderline symptomatology. In conclusion, the DERS‐32 has demonstrated to be the best model for the DERS among all the others considered in this work, as well as a reliable tool to assess deficits in ER.