the degree to which the values predicted by a model—typically, one developed in regression analysis, structural equation modeling, or confirmatory factor analysis—diverge from the corresponding empirical values. A large, significant value from a goodness-of-fit test indicates a poorly fitting model.