Predicting Statistics Achievement: A Prototypical Regression Analysis

Abstract

The purposes of the current study are: (a) to demonstrate a viable approach to the conduct of a multiple regression/correlation analysis; and (b) to illustrate the approach in the context of predicting achievement in an introductory statistical methods course. The analysis is proposed as being appropriate if the basic intent of a study is that of prediction as opposed to that of explanation. That is, the intent is to arrive at a model for predicting a criterion in as efficient a manner as the data on hand will allow. No model, causal or otherwise, is being posited or verified.