Thorndike, on the other hand, in studying the relation between entrance marks and later college standing (Columbia College classes entering in 1901, 1902 and 1903), finds results which lead him to say, "The important facts concerning the relationship of success in entrance examinations to success in college work... prove that we cannot estimate the latter from the former with enough accuracy to make the entrance examinations worth taking or to prevent gross and intolerable injustice being done to many individuals.... The records of eleven entrance examinations give a less accurate prophecy of what a student will do in the latter half of his college course than does the college record of his brother! The correlation between brothers in intellectual ability is approximately .40, but that between standing in entrance examinations and standing in college of the same person is only .47 for junior year (130 cases) and .25 for senior year (56 cases).... From many facts such as these... it is certain that the traditional entrance examinations, even when as fully safeguarded as in the case of those given by the College Entrance Examination Board, do not prevent incompetence from getting into college; do not prevent students of excellent promise from being discouraged or barred out altogether; do not measure fitness for college well enough to earn the respect of students or teachers; and do intolerable injustice to individuals."
The apparent striking contradiction between these two reports is not, however, so serious when it is noted that the records of Jones were taken from freshman and sophomore years, while Thorndike's, as here quoted, were taken from junior and senior years. Thorndike has also presented, in another connection, comparisons of entrance examinations with the work of freshman and sophomore years, and in these cases his correlations are considerably higher, more nearly approximating the results of Jones. The correlations, for the four college years, were as follows: freshman year, .62; sophomore year, .50; junior year, .47; senior year, .25.
Apparently the only safe conclusion at present is that the entrance examinations are fairly useful in predicting the early college work, their prognostic value becoming less and less as the interval between the two measures is increased. This result is of course to be expected. In another section of this book occasion is taken to show that preliminary trials are of little value in indicating the relative abilities of individuals when they have reached or approximated their limit of practice.