An Aspect of Logistic Regression : Goodness of fit tests for the Logistic Regression Model

This text is a preliminary look at logistic regression model-building with respect to determine whether the constructed regression model accurately describes the data, whether the regression model significantly describes the outcome variable given the explanatory variables. A brief background of logistic regression is given, including the mathematical derivations for the regression model. A look at various goodness-of-fit tests in the development of logistic regression models with an accompanied example using Matlab to examine the strengths and weaknesses of goodness-of-fit statistics. The text compares goodness-of-fit statistics as to which is appropriate given the conditions of the data, and examines the strengths and weakness of these statistics. The Matlab programs and codes used for the examples in the text are included in the appendix, along with all necessary tables.