Application of Local Influence Diagnostics to the Linear Logistic Regression Models

This paper focuses the development of the diagnostics for the perturbations of case-weights and explanatory variables (one or more) in a linear logistic regression model. The effect of specific perturbation scheme on the estimation of parameters is also assessed. In addition, the interpretation of the value of curvature diagnostics is highlighted in this paper. This paper also demonstrates and extends the utility of the diagnostics for dichotomous outcome variables. For illustration, a sub-set of the Framingham Heart Study data set is used.