Presents two viewpoints about the use and interpretability of beta weights in educational research: (1) that beta weights should be interpreted as a logical index of the importance of individual predictors within the context of the entire set of predictors; and (2) that interpretation requires certain cautions and conditions. (SV)

This article illustrates the role multiple linear regression can play in developing prediction equations by providing examples of regression models that could be used in answering questions relative to the importance of a single predictor variable, interactions between predictor variables, and the cross-validation and generalizability of…

The utility of a nonlinear transformation of the criterion variable in multiple regression analysis is established. A well-known law--the Pythagorean Theorem--illustrates the point. (Author/JKS)

The utility of a non-linear transformation of the criterion is established. The Pythagorean Theorem is used as the example to demonstrate the point. The functional relationships may be such (as in the Pythagorean Theorem) that an R-squared of 1.00 cannot be found without making a non-linear transformation of the criterion. The goal of…