Differential Prediction of Ability as Represented by College Subject Groups

Abstract

The problem of predicting scholastic success has been the subject of considerable thought and investigation since the advent of measurement in the field of education. The problem is now more pertinent than ever owing to the rapid development of public education through the enforce ment of compulsory education and the upper extension of education in the junior college. This change in public education is in line with our recognition of our responsibility for creating the opportunity for all children to develop to the limits of their ability. The prediction of scholastic success with the use of tests and meas urements in education began with the correlation of marks made by pupils one year with the marks made in later years, as was done by Miles.1 The advance was made by Kelley2 through the use of the multiple re gression equation. At first, teachers' marks were used and then later the results of tests and questionnaires were added as items for predic tion. All such correlations and predictive equations have been directed towards two sorts of criteria. The one criterion of scholastic success has been the general average of a student's success in some one year or in an institution. The other criterion has been the success in some specific subject or group of subjects. On account of the tendencies in the growth of education noted above, it seems desirable to add a new criterion of scholastic success to the two already listed. This criterion is the differential prediction of scholastic success. If all children must attend school for longer and longer periods of time it becomes more and more important that we know in what sub jects any specific child is most likely to be successful. The prediction of general scholastic success is not always so important since children must attend school in any event. Once the fundamentals of education