While research has shown the statistical significance of high school grade point averages (HSGPAs) in predicting future academic outcomes, the systems with which HSGPAs are calculated vary drastically across schools. Some schools employ unweighted grades that carry the same point value regardless of the course in which they are earned; other…

In 2002, an article entitled "Four assumptions of multiple regression that researchers should always test" by Osborne and Waters was published in "PARE." This article has gone on to be viewed more than 275,000 times (as of August 2013), and it is one of the first results displayed in a Google search for "regression…

In a multiple regression analysis with three or more predictors, every set of alternate weights belongs to an infinite class of "fungible weights" (Waller, Psychometrica, "in press") that yields identical "SSE" (sum of squared errors) and R[superscript 2] values. When the R[superscript 2] using the alternate weights is a fixed value, fungible…

Using self-reported academic cheating from the classes of 1959 through 2002 at the three major United States military service academies (Air Force, Army, and Navy), we measure how peer cheating influences individual cheating behavior. We find higher levels of peer cheating result in a substantially increased probability that an individual will…

The cross-validated classification accuracies of three predictor weighting strategies (least squares, ridge regression, and reduced rank) were compared under varying simulated data conditions for the two-group classification problem. Results were somewhat similar to previous findings with multiple regression when absolute rather than relative…

For idealized item configurations, equal item weights are often virtually as good for a particular predictive purpose as the item weights that are theoretically optimal. What has not been clear, however, is what happens to the similarity when the item configuration's variance structure is complex. (Author/CTM)

Factor scale scores are sometimes used as weights to create composite variables representing the variables included in a factor analysis. If these composite variables are then used to predict some dependent variable, serious theoretical and methodological problems arise. This paper explores these problems and suggests strategies for circumventing…

Four methods were used to weight predictors associated with a Resident Assistant job: (1) rank order weights; (2) unit weights; (3) critical incident weights; and (4) regression weights. A cross-validation was also done. Most weighting methods were highly related. No method was superior in terms of protection from validity shrinkage. (GDC)

Computer-based Monte Carlo methods compared the predictive accuracy upon replication of regression of five complete and four incomplete factor score estimation methods. Prediction on incomplete factor scores showed better double cross-validated prediction accuracy than on complete scores. The unique unit-weighted factor score was superior among…

The relative magnitudes of R-squared values computed through multiple regression models using grade equivalent scores, raw scores, standard scores, and percentiles as both predictor and criterion variables are compared. Grade equivalents and standard scores produced the highest R-squared values. (Author/JKS)