A Psychometric Analysis of the Duke Misophonia Questionnaire.

Abstract

Background: The Duke Misophonia Questionnaire (DMQ) is one of few psychometrically constructed scales designed to measure misophonia, a condition characterized by heightened sensitivity to specific auditory stimuli, resulting in strong emotional, physiological, and behavioral responses. While prior research has evaluated the dimensionality of individual subscales, a comprehensive psychometric analysis of the entire DMQ has not been conducted.

Objectives: This study aimed to extend the psychometric evaluation of the DMQ using contemporary psychometric methodologies to enhance its consistency, stability, and applicability in future research.

Methods: Two samples of university students were recruited online via a research participation pool (Sample 1: N = 318; Sample 2: N = 424; both predominantly female and White, aged 18-30). A novel, empirically derived five-factor model (beliefs, coping, impairment, aversion, and anxiety) and the theoretically defined eight-factor model, developed from clinical and theoretical foundations, were assessed. The five-factor model was further examined using multidimensional item response theory (MIRT) to gain insights into item- and test-level functioning.

Results: Both models demonstrated adequate fit, though the five-factor model was more parsimonious and fit slightly better. Neither model contained cross-loadings. The five-factor model was further evaluated through MIRT which found that the DMQ provides the most information at moderate to high levels of misophonia severity. This analysis provided valuable information regarding the functioning and efficiency of the DMQ at both item and test levels.

Conclusion: The results contribute to the refinement of the DMQ by identifying a more parsimonious psychometric structure, promoting greater consistency and stability in its use for future misophonia research.