Related Experiment Videos
Examining the construct validity of the Eating Disorder Inventory
Dorothy L Espelage1, Suzanne E Mazzeo, Steven H Aggen
1Department of Educational Psychology, University of Illinois at Urbana-Champaign, 226 Education Building, 1310 South Sixth Street, Champaign, Illinois 61820-6990, USA. espelage@uiuc.edu
Psychological Assessment
|April 5, 2003
Summary
The Eating Disorder Inventory's (EDI) construct validity was evaluated. A two-factor model (Eating and Weight, Personality) best fit the data, supporting its use in assessing eating disorders.
Area of Science:
- Psychometrics
- Clinical Psychology
- Eating Disorder Research
Background:
- The Eating Disorder Inventory (EDI) is a widely used self-report measure for assessing symptoms associated with eating disorders.
- Examining the construct validity of the EDI is crucial for ensuring its accurate interpretation and application in clinical and research settings.
Purpose of the Study:
- To evaluate the construct validity of the Eating Disorder Inventory (EDI) across multiple samples.
- To determine the optimal factor structure of the EDI and assess its reliability and convergent validity.
Main Methods:
- Confirmatory factor analyses (CFAs) were conducted on item-level and item-parcel data from three distinct samples (N=855).
- The study utilized the EDI, a structured interview, and the Millon Clinical Multiaxial Inventory-II (MCMI-II) for data collection.
- Reliability estimates and interscale correlations were calculated to assess scale consistency and convergent validity.
Main Results:
- Neither null nor one-factor models adequately fit the EDI data.
- The proposed eight-factor model showed a good fit to item-parcel data, but not item-level data.
- A final CFA supported a two-factor model (Eating and Weight, Personality) for the eight EDI scales. High reliability was found for the scales (0.82–0.93).
- EDI personality scales demonstrated associations with MCMI-II scales, and significant correlations were observed between depression and EDI scores in a treatment sample.
Conclusions:
- The findings partially support the construct validity of the EDI, with a two-factor structure providing the best fit.
- The EDI demonstrates adequate reliability and convergent validity, suggesting its utility in differentiating individuals with and without eating disorders.
- Further research may refine the EDI's factor structure and its application in diverse clinical populations.