Expanding the Use of Weight of Evidence and Information Value to Continuous Dependent Variables for Variable Reduction and Scorecard Development

Weight of Evidence (WOE) and Information Value (IV) have become important tools for analyzing and modeling
binary outcomes such as default in payment, response to a marketing campaign, etc. By suggesting a set of
alternative formulae, this paper attempts to expand their use to continuous outcomes such as sales volume, claim
amount, etc. A SAS® macro program is provided that will efficiently evaluate the predictive power of continuous,
ordinal and categorical variables together and yield useful suggestions for variable reduction/selection and for
segmentation. Modelers can use the insights conveyed by SAS outputs for subsequent linear regression, logistic
regression or scorecard development.