Date of Award:


Document Type:


Degree Name:

Master of Science (MS)



Committee Chair(s)

David M. Stein


David M. Stein


Gretchen Gimpel


Kevin Masters


The Anorexia Bulimia Inventory, a recently developed self-report questionnaire for the assessment of eating disorders, addresses two major limitations found in existing self-report eating disorder inventories. First, it comprehensively assesses the diagnostic symptoms of both bulimia and anorexia nervosa; and second, it assesses the frequently cooccurring problem areas (e.g., depression, anergia) that may be targeted in treatment planning for eating disorders. Although initial research on the psychometrics of the instrument appears promising, no research has yet investigated its factor structure. Therefore, the goal of the present study was to investigate the factor structure of the Anorexia Bulimia Inventory.

Principal axis factor analysis with a varimax rotation was used on a combined clinical and nonclinical sample. The overall sample (N = 1,675) was randomly divided to provide a replication of the factor analysis. Using the two separate samples and the combined, total sample, one seven-factor solution and two eight-factor solutions emerged. All factors met conventional standards for internal consistency, with the exception of one factor consisting of only three items.

The results of this study suggest that the factor structure of the Anorexia Bulimia Inventory closely parallels the intuitively designed subscales. All nine subscales emerged as separate, interpretable factors in at least one sample. Four subscales were clearly corroborated by the factor analysis, suggesting that these factors represent stable constructs that are relevant to populations at-risk for the development of eating disorders. The factor analyses provided moderate corroboration of four other subscales. Although these four factors did show deviations across samples, they all emerged as interpretable factors, suggesting that the constructs likely reflect some of the prominent ideational and behavioral issues relevant to at-risk populations. Finally, one subscale emerged as an interpretable factor in only one sample. This failure to replicate across samples may reflect problems in the wording or choice of items included on the subscale, or it could indicate that these issues are less relevant for at-risk populations. The results of this study, together with the existing evidence of the measure's psychometric properties, support the use of the Anorexia Bulimia Inventory in the assessment of eating disorders.



Included in

Psychology Commons