The reliability, stability, and validity of an Implicit Association Test for body image

The Implicit Association Test (IAT) was adapted to measure body image by assessing association of the body (my body vs. not my body) with unpleasant or pleasant words on eighty-four female subjects. This exploratory study was designed to answer three general questions: Could the IAT be adapted to study body image, would it be internally reliable and stable across time, and would it display divergent and predictive validity? The study demonstrated that the IAT-Body Image (IAT-BI) exhibited good internal consistency and adequate stability. Confirmatory factor analysis (CFA) showed that the IAT measure defined a factor that was distinct from, but weakly correlated with, a factor defined by standard explicit (self-report) body image measures. However, the IAT-BI failed to demonstrate acceptable predictive validity. Instead, the data demonstrated that explicit measures were superior predictors of self-reported and observed criterion behaviors.