Columbia University by the examination of the College Entrance Examination Board of the Middle States and Maryland. The examinations of this board are universally admitted to be the most efficiently administered of any in this country; and their results are accepted at their face value by every college and university in the United States. So Professor Thorndike's study may be conceded to have been a test of the examination method at its best. His significant conclusions are given herewith, in his own words. The text of his article may be found in full in the Educational Review for May, 1906.

Professor Thorndike says:

The important facts concerning the relationship of success in entrance examinations to success in college work are given in the tables at the end of this article. They 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. . .

...

Of the students who were in the lower half of the group in the entrance examinations nearly 40 per cent are found in the upper half in the last three years of college.

Of the dozen students who ranked highest in entrance, some were in the lowest fifty of the class by the junior year.

If, knowing that 50 individuals ranked in the order Jones, Smith, Brown, etc., in their entrance marks, one were to wager that in the college work of, say, junior year, they would rank Jones, Smith, Brown, etc., as before, he would lose his bet in 47 cases out of the 50.

...

The records of eleven or more entrance examinations gives 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. . . . It is a moral atrocity to decide the fitness of an individual for college by a system which, when required to work to a moderate degree of accuracy, is wrong 47 times out of 50! ...

And then, by way of comment, he adds:

A sagacious tutor can get a hundred boys into college [by examination], not one of whom he would be willing to certify as fit to succeed there.

Below will be given an account of the scholarship records of those students who enter the University of California by a combination of recommendation and examination, in comparison with the records of those who enter wholly by recommendation. The results will be seen to be quite in accord with Professor Thorndike's conclusions as given above.

There is still another objection to the present mixed method of admitting students to the University. For several years now the writer has made a careful study of the scholarship records made in the University by all the freshmen from accredited schools. The results of this study are used as one element in determining the elegibility of the school for accrediting the following year. Furthermore, these records are published yearly, with the President's Report, and a copy is sent to the principal of every secondary school in the state. It is evident that this wide publicity must act as a strong stimulus upon the schools. Naturally, each school desires to maintain a record for scholarship on the part of its representatives which will bear favorable comparison with the records of the other secondary schools of the state.

Now, as noted above, it frequently happens, with our present method of admission, that a student who brings fewer than the full required number of "recommended" units for admission supplements these credits by examination taken at the University, and so gains admission, which, so far as the school was concerned, was denied him. Now it is manifestly unfair to charge the record of such a student to the school from which he came. If the records of students so entering were as good as those of the fully recommended students, no practical harm would result, and probably no one would seriously object. This, however, is by no means the case, as will be shown later.

Thus the University finds itself in the unenviable position of putting upon the schools the responsibility for low records in scholarship for which it is itself really respon

sible, and thus transferring to the schools its own shortcomings. The injustice, not to mention lack of dignity, of such a procedure must be self-evident; the schools fully recognize and frequently complain of this fact.

This brings us to the fourth, and, from the University point of view, the most weighty, of all the objections to the present system.

The school certifies to satisfactory work, when such has been done; refusal to "recommend" naturally means that, in the judgment of the school, the work of the applicant in the subject in question was not sufficiently well done to warrant its being used for college entrance. Actually, then, every time the University examines an applicant graduated from an accredited school, in a subject studied in the school, its action can admit of but one interpretation: it is a thinly veiled, even if unintended, suggestion on the part of the University that the standards of the school are too high. So we have this curious spectacle: a great university that has been ostensibly trying for more than a generation to elevate the standards of the schools and the scholarship of the freshmen intrants is found actually, if not intentionally, debasing the standards of both.

That the standards of the schools do tend thus to be actually lowered may be only a natural though unprovable inference; but in considering the effects of this policy upon the standards of admission, we have a solid basis of fact upon which to rest our case. The writer has critically studied the data relating to all those freshmen from California schools who entered the University in 1912-13 and took any examination whatever to supplement the partial recommendations which they received from the schools. It will be noted that, since they did not have the requisite number of "recommended" units, their final admission was actually due to the action of the University itself.

Eighty such students entered from public high schools, and five from private schools; for purposes of simplification, the cases of the latter are not considered in what follows. Their cases, however, form no exception to the rule.

The general scholarship index of these eighty students made in the first term of their University residence was 2.87 -a strikingly low record, as compared with 2.43, which was the corresponding index for the fully recommended students for the same time.

Still more striking and significant, for our present purpose, is the distribution of the work of these eightly partly "examined" students in the various University grades, as compared with the distribution of the work of the fully "recommended" students.

The following table shows the percentage of work done in the different University grades:

From this it appears that the "examined" intrants made 23 per cent less of first and second grade work and 36 per cent more of fourth and fifth grade work than did the fully "recommended" students.

In the light of these facts, with what grace can the University complain of poorly prepared freshmen ?

Now, a simple change in the present method of procedure in reference to the admission of students from accredited schools of the state from which more than ninety per cent of the freshmen come would, the writer believes, not only remove all these and like evils, but would work a very positive good besides.

It is proposed that, beginning at some definite date, say, August, 1914, the University should, after due notice had been given, decline to consider for freshman admission any graduate of an accredited California secondary school who is under twenty-one years of age, except he bring an unqualified recommendation for such admission from the school from which he was graduated. Details concerning the form of this recommendation are suggested below.

Should such a step be taken, it would mean that the University had taken its stand, both in theory and in practice, upon the logical implication of the accrediting system, viz., that the school which has trained the pupil for his college work is the only proper judge of his fitness to go to college, and that the school should, accordingly, be held wholly responsible for the decision of this matter.

This change in method would involve but slight modification of actual practice, as pointed out above, present practice actually accords with it in the care of all fully "recommended" students, and they constitute more than ninety per cent of all freshmen intrants from California.

A further examination of freshmen admission data for the year 1912-13 shows that the eighty-five students referred to above as having entered partly by recommendation and partly by examination were but eight and one-half per cent of the total freshmen admissions for the year from California schools. The total amount of "recommended" work presented by the 923 freshmen who entered wholly without examination was 44,450 units; the eighty-five partly examined students undertook a total of only 626.5 units of examination.

Thus it appears that the examined units were but 1.4 per cent of the total freshman admission units for the year. The units offered by the fully recommended applicants averaged 48.15 per person; the examinations averaged but 7.4 units each.

These "examined" units were distributed as follows: out of a total entrance requirement of forty-five units for each applicant, 38.8 per cent of the applicants took examinations in only three units each; 22.3 per cent of them took examinations in only six units each; the remaining 39 per cent were scattered between seven and one-half and twentyseven units each, with but a few representatives for each number. The proportion of "examined" units passed was 84 per cent.