Another Look at the Method of Y-Standardization in Logit and Probit Models

摘要: This article takes another look at the derivation of method Y-standardization used in sociological analyses involving comparisons coefficients across nested logit or probit models. It shows that can be derived under less restrictive assumptions than hitherto suggested. Rather assuming fixes variance latent error a known constant, it suffices to assume is unknown. A further result suggests using for cross-model likely biased by model differences fit assumed logistic normal distribution. Under correct specification recovers an effect size metric similar Cohen's d.