Is the widely used two-factor structure of the Ruminative Responses Scale invariant across different samples of women?

Abstract

Although the Ruminative Responses Scale is one of the most widely used measures of rumination, its two-factor structure remains controversial. Taking this into account, we aimed to test the RRS-10 two-factor invariance (Brazilian version) between different samples of women and to study its internal consistency and convergent validity. A sample of 321 women (general population, n=106; college students, n=115; and medical population of patients with overweight and obesity, n=100) participated in the study. The two-factor structure of RRS-10 was assessed by CFA and multigroup analysis using Mplus software. Internal consistency was assessed by Cronbach's alpha and the convergent validity by Pearson correlations. The two-factor structure of RRS-10 showed a good fit, factorial invariance across three samples, good internal consistency, and adequate convergent validity. Brooding and Reflection subscales were both positively correlated with psychological inflexibility, cognitive fusion, anxiety, depression, and stress symptoms, although Brooding presented significantly stronger associations with these variables than Reflection. This study provides further discussion and evidence regarding the RRS-10 two-factor structure, as well as a valid version of RRS-10 to use in Brazil in order to reliably assess rumination in medical and research settings. This is the first study to test and confirm the RRS two-factor structure invariance across groups. RRS-10 two-dimensionality was confirmed in medical and non-medical samples of women. Brooding subscale showed significantly stronger relationships with psychopathology and experiential avoidance than Reflection. The study provides evidence that RRS can be used as a valid and sound measure to accurately assess the clinically relevant dimensions of rumination simultaneously across distinct groups.