Logistic discriminant analysis is frequently applied to data sets with discrete observed variables. When atypical observations exist in a data set, they may exert undue influence on the result of the analysis. The local influence approach is employed to develop diagnostic measures for...