skip to main content skip to footer

Preliminary Report on the Prediction of Academic Performance at the California Institute of Technology

Author(s):
Educational Testing Service
Publication Year:
1949
Report Number:
RB-49-04
Source:
ETS Research Bulletin
Document Type:
Report
Page Count:
11
Subject/Key Words:
California Institute of Technology, College Board Achievement Tests, College Entrance Examinations, Mathematics Achievement Test (College Board), Grade Prediction, Predictive Validity

Abstract

This study was undertaken to determine the effectiveness of the College Entrance Examination Board tests in predicting academic performance at the California Institute of Technology and to compare the predictive power of these tests with the entrance examinations presently being used. The results of this investigation indicate that the College Board tests have satisfactory validity as predictors of success at the California Institute of Technology. Together they yield a multiple correlation with first term Grade Point Average of .67 and individually they correlate with first term grades in corresponding courses for .48 to .59. These results compare very favorably with results of similar studies at other institutions. Comparison of the predictive efficiency of the CEEB tests and the CIT entrance examinations is made difficult because the students were selected on the latter. The resulting restriction in range, being greater on the CIT tests than on the CEEB, vitiates a direct comparison which can only be made if distributions of scores of all applicants on the CIT tests can be obtained. On the basis of the evidence at hand, it seems likely that the only test correlations markedly affected by selection are those of CIT mathematics with Mathematics grade and Grade Point Average. Tentatively, it appears that the CEEB Mathematics, Chemistry, and English (Verbal) tests are more effective than the comparable CIT tests while the latter's Physics test is more predictive of success at CIT than the CEEB's.

Read More