ABSTRACT
Determining the most important predictor variables in a regression model is a vital element in the interpretation of the model. A general rule is to view the predictor variable with the largest standardized coefficient (SRC) as the most important variable, the predictor variable with the next largest SRC as the next important variable, and so on. This chapter presents an alternative measure—the predictive contribution coefficient—that offers greater utile information than the SRC as it is an assumption-free measure founded in the data mining paradigm. The proposed Predictive Contribution Coefficient (PCC) is a development in the data mining paradigm. The PCC is flexible because it is an assumption-free measure that works equally well with ordinary and logistic regression models. The PCC decile-based small data calculations, which are obvious and trivial, are presented to make the PCC concept and procedure clear and to generate interest in its application.
