A bifactor approach to modeling the structure of repetitive thinking: Replication and extension

Recent research found that a bifactor structure provides the best fit for maladaptive repetitive thinking. The present research attempted to: (a) replicate this finding with an alternative measure of rumination; and (b) to extend it by testing the best fitting model including both adaptive and maladaptive repetitive thinking using the Penn State Worry Questionnaire and the Rumination and Reflection Questionnaire. Using data from 14 independent datasets (N = 4711), the fit of three worry and rumination models and five worry, rumination, and reflection models were compared. The validity of the latent factors of the best fitting model of worry, rumination, and reflection was examined with several factor integrity tests and by examining their associations with theoretically relevant constructs. Results indicate that a bifactor structure provides the best fit for worry, rumination, and reflection. Furthermore, four orthogonal factors derived from the bifactor model showed diverging associations with distress, personality, emotion regulation, emotional awareness, and intolerance of uncertainty.