Prediction of second-year and third-year grade-point averages at the U.S. Naval Postgraduate School

Abstract

Further validation of the 20 aptitude and achievement tests given in 1948 and of the 12 tests given to the group entering the Postgraduate School in 1949 has been carried out. The 1948 test scores have been related to quality-point ratios for the entire second year and for the first two terms of the third year, both by bar charts depicting the per cent of students in each third of the score range on each test who achieved a 2.0 quality-point ratio or above, and also by means of product-moment coefficients of correlation. For the 1948 tests, mean test scores of drop-outs from the School were also compared with mean scores for those continuing in Postgraduate work. Similarly bar charts were prepared and product-moment coefficients of correlation were computed for the 1949 tests. While the phenomenon of shrinkage in test validities from first to second year previously observed with the 1948 tests also was found with the 1949 tests, the shrinkage was not as great. Also, the validities of 1948 tests against third-year grade-point ratios tended to be somewhat higher than those for the second year. It was concluded on the basis of the findings that tests of advanced mathematics, mathematics aptitude, physics, and engineering are effective for predicting success at the Postgraduate School as judged by all three of the techniques employed for assessing the tests.