Predictor and Performance Variables in a Performance-Based Education Course

Abstract

Students enrolled in classes that stressed learning theory and human development served as subjects in an investigation of the predictive power of observation scores compared with more traditional predictors such as grade point average and American College Test scores. A stepwise discriminant analysis was employed, using the BMDP7B program on the Sperry 1100 Univac, to distinguish the relative predictive strength of the selected variables on the variable of achievement reflected by grade. In the first step of the analysis, observation scores were significantly above the other variables, confirming our basic hypothesis. The second step in the analysis revealed significance of the observation scores and the college grade point average scores. The results suggest that learning that occurs through observation of selected principles compares more than favorably with standardized testing and grade point averages in predicting success in field-based learning.