Objective: Disordered eating attitudes and behaviors play a critical role in pediatric overweight and obesity and are important to measure in research and practice with this population. Disordered eating attitudes and behaviors in youth are commonly measured using the Children's Eating Attitudes Test (ChEAT), which has not been psychometrically evaluated for use with youth with overweight and obesity. As such, it currently unknown whether the ChEAT provides a reliable measurement of a single underlying construct when applied in this population. Methods: Data from 400 treatment-seeking youth (aged 8 to 18) with overweight and obesity were drawn from two sites in the southern United States (54% female, 57% White). Psychometric properties of the ChEAT were evaluated in this combined sample using confirmatory bifactor modeling. Results: A confirmatory bifactor model with five group factors provided the best fit to the data and was used to generate numerous psychometric coefficients. The explained common variance (.36), omega hierarchical (.58), factor determinacy (.86), and construct replicability (.70) coefficients were all lower than recommended cutoff criteria. Both multidimensionality and excessive residual variance contributed to the ChEAT's poor psychometric performance in this sample. Conclusions: The ChEAT displayed marked psychometric problems in a large and diverse sample of youth with overweight and obesity. Its common variance was not unidimensional, its unit-weighted total score was not a reliable indicator of a broad central construct, and its optimally weighted total score was not likely to replicate across studies. We recommend using alternatives to the ChEAT when conducting research and practice with youth with overweight and obesity.
Psychometric analysis of the Children's Eating Attitudes Test in youth with overweight/obesity
