The Modified Weight Bias Internalization Scale: measurement invariance by weight status and race among undergraduate women.

Background: Internalized weight bias is the belief in negative, weight-based stereotypes and the application of these stereotypes to oneself. These negative stereotypes have harmful impacts on people with overweight/obesity, and weight-based discrimination is well-documented across a variety of settings. Given poor outcomes associated with internalized weight bias, particularly among individuals with obesity, it is necessary to validate measures assessing internalized weight bias among diverse samples. The present study sets out to investigate measurement invariance properties across weight status (women with vs. without overweight/obesity) and race (White vs. Asian; White vs. bi- or multi-racial) of the Modified Weight Bias Internalization Scale (WBIS-M), an 11 item self-report measure.

Methods: Participants were 746 racially/ethnically diverse women across the weight spectrum (24.9% with overweight/obesity). Confirmatory factor analyses of the WBIS-M were initially performed among the full sample, and all sub-samples. Each model showed good to excellent descriptive model fit. Subsequent analyses examined factor loadings and item thresholds of the WBIS-M to assess metric, threshold, and scalar invariance. Invariance was determined by assessing changes in Comparative Fit Index (ΔCFI $\le$ -0.010), Root Mean Square Error of Approximation (ΔRMSEA $\le$ 0.015), and Standardized Root Mean Square Residuals (ΔSRMR $\le$ 0.030).

Results: Based on these previously established statistical cutoffs, the WBIS-M showed invariance across weight status and racial groups in the present sample. The current results lend support for use of the WBIS-M to measure internalized weight bias in women who do and do not have overweight/obesity, and among White, Asian, and bi- or multi-racial women.

Conclusion: This may inform future studies that wish to utilize the WBIS-M, such as investigations of mean level differences in internalized weight bias. These findings may have clinical applications in the treatment and prevention of obesity, given the heightened levels of internalized weight bias and weight-based discrimination faced by individuals with higher body weights.