Children’s Eating Attitudes Test: Revised factor structure for adolescent girls

Abstract

The Eating Attitude Test (EAT) and its language-simplified version for children (ChEAT) have been popular instruments for the assessment of eating attitudes among children, adolescents and young adult females. Data collected from female adolescents using the ChEAT were analyzed to compare five previously proposed measurement models as well as a new model. Confirmatory factor analysis (CFA) was used to directly compare models in terms of their goodness-of-fit indices and to determine which model best fits the overall data (N=737) as well as data from the youngest and oldest girls (N=200 each). A newly proposed 14-item, 5-factor model provided the best fit to the overall data as well as data analyzed separately for the youngest and oldest girls. Furthermore, each of the newly proposed factor scores varied somewhat independently as a function of grade level, with body image concerns increasing most dramatically from grades 5 through 8. This newly proposed model is based on factors previously suggested by others, but not previously combined into a single measurement model. Implications of the results for future research involving adolescent girls are discussed, especially research on developmental changes in eating attitudes and behaviors that may constitute risk factors for subsequent eating disorders.