A Hierarchical Cluster Analysis of the Core Courses in an Engineering Curriculum

THE PROBLEM of selecting students for univer sity and college admittance is pervasive and ever present and considerable attention has been devoted to developing effective selection procedures. How ever, most research efforts in this area have con centrated on the predictor segment of the prediction equation and have paid very little attention to the cri terion side. Most often the criterion used for eval uating the worth of a selection instrument is fir s t term or first year total grade point average and not much effort has been devoted to examining what di mensions or ''factors" might explain the total vari ance of this global type measure for various cur ricula. The present study attempts, by means of a step wise hierarchical cluster analysis, to break down this global measure in the most meaningful manner possible so as to provide information relevant to what factors of course content account for the vari ation in the total GPA, which predictors in the bat tery predict which course clusters best, and what content factors need better predictors to account for their variance.