Factor Structure and Psychometric Properties of the ED‐15 in People With an Eating Disorder

Abstract

ABSTRACTObjectiveThis study examines the factorial structure and psychometric properties of the Eating Disorder‐15 questionnaire (ED‐15) in a large clinical sample, as well as the instrument's sensitivity to early clinical change in therapy and ability to measure remission.MethodParticipants with eating disorders (N = 278) referred to the Flinders University Services for Eating Disorders in South Australia completed the ED‐15 as well as other measures of eating disorder symptoms and co‐occurring psychopathology, including depression, anxiety, and stress.ResultsConfirmatory factor analysis (CFA) revealed a two‐factor model for the ED‐15. The ED‐15 had good internal consistency. It showed satisfactory concurrent validity with moderate correlations with the EDE‐Q global score and contribution of unique variance to that score. Correlations indicated good convergent validity with clinical impairment and good divergent validity from depression, anxiety, and stress. The ED‐15 showed a significant medium effect size change within the first four sessions of therapy. Good discriminant validity was indicated by cut‐off scores used for remission, with significantly different levels of ED psychopathology and other impairments between the two groups.DiscussionThis study adds to the four previous psychometric studies of the ED‐15, confirming robustness of the English version in a clinical sample. The brevity and psychometric robustness of the ED‐15 makes it a preferable measure to the Eating Disorder Examination for sessional assessment of progress in treatment.