SUCCESSFUL universities have always remained mercifully vague about the purposes of education; national programs to improve education have perpetually sought to define the aims of education. For this reason, the Scholastic Aptitude Test, which was a boon to overworked admissions officers also threatened to create a relatively narrow definition of excellence; it was the attraction of simplification and the hazard of limiting diversity by over-emphasizing test scores that led a former Dean of Admissions to warn against "the tyranny of little numbers."
One barrier against such tyranny, at Harvard and elsewhere, has been the relative inaccuracy with which present tests predict college performance. The National Merit Scholarship Corporation, in an effort to find tests of greater "usefulness," is developing criteria for admission that promise to predict college grades with unprecedented accuracy by combining the familiar Scholastic Aptitude Test with several standard personality tests. Already past the first stages of research, the Merit Corporation is now investigating the different personality factors associated with performance in individual colleges.
And the new investigations are disturbing, because Merit, which has done most of the research, has a very doubtful judgment about the purposes of college. To measure the "productivity" of a college, for example, they rely on the so-called Knapp-Greenbaum index: the percentage of graduates of that college who go on to get a Ph.D. The Merit people apparently had some reservations about this index--not, however, because they think that colleges might "produce" something other than scholars, nor because they think it might be unwise for a college to lure all its students into scholarship.
They simply recognized that the Knapp-Greenbaum index did not distinguish between the effects of the college and the quality of the students who came to the college; and so, confident that the qualities of a college could be separated from the quality of the student body, an NMSC worthy named Thistlethwaite devised the Talent Supply Index, which measures the calibre of students--a college's TSI is the average freshman's score on the Scholastic Aptitude Test.
And Merit is very fond of test scores. Despite the vast quantities of evaluation that schools supply about National Merit Scholarship candidates, the scores seem to be crucial. The Center for the Study of Higher Education administered some personality tests to students after National Merit made its scholarship awards, and discovered that while a criterion called Complexity of Outlook distinguished winning girls from the runners-up, Scholastic Aptitude scores were the only other significant difference between either male or female winners and those who did not win.
RECENTLY the National Science Foundation conducted a study of the way that awards were made, and discovered, again, that judges who had both test scores and descriptions of the candidate tended to rely on the scores.
You need not believe that Harvard is consciously directing its students into graduate school, nor that it is solely concerned with grades, to wonder at the effects of new scoring systems that will predict both with great accuracy. The temptation to rely on scores, as Merit and the National Science Foundation have shown, can overpower good intentions, and though today Harvard may be little concerned that students seem to attend graduate school with increasing frequency, it may be forced to take a more
For it would seem almost impossible not to consider the kind service these graduates will
Perhaps it is time, also, to
If National Merit can truly