measure diligence in work, knowledge of facts and the grasp of principles is the most difficult and the most important single act any college instructor is called upon to perform; that is, to measure the three. It is much easier to measure for any one than more. It requires a knowledge not only of the subject, but of the instructor himself and of the capacity and the habits of his pupils. The examination, at best, measures the teaching and the teacher himself, as well as the students to be taught. If you want to know how efficient the instruction in a course is, read the examination books after they have been marked, and that tells you. If you want a standard by which to measure yourselves, read the examination books in a course as nearly as possible parallel to your own. I remember very well, before being an instructor here, when I was on a visiting committee, one of the instructors gave me that suggestion. Some of the committee went to two or three lectures in various courses, and then we asked the instructor to send us books. This was the suggestion of a former instructor in the department, a professor emeritus. We sent for a number of books. We said, "Send us a couple of A; couple of B; couple of C; couple of D books." We did that in each of the various courses we visited. We read them, and we got an extraordinarily clear impression of the instructor. We were not after the pupils, and we got no particular impression of them, but we got an extraordinary impression of the instructor. It so happened that two of the courses were virtually parallel. I won't mention what they were, as you might discover who the people were, although they have all disappeared from sight, and I believe them dead. In one of them, it was perfectly evident that the instructor had given his pupils a grasp of the meaning of the things he was talking about. The other man, curiously enough, happened to ask nearly the same questions, and it was perfectly evident that none of his pupils had anything but a rote knowledge. They did not understand the underlying principles. It was astonishing, the mental picture we got of the instructor from reading those ten books. We got an extraordinarily exact idea of the instructor's capacity and his power of imparting..

If we are earnestly seeking for a measure of efficiency, means can certainly be devised, whether by a general examination upon

the subject at the end of a period of instruction covering many courses, or by submitting examination books from time to time to colleagues in the same or another institution, or by some other test it would certainly be possible for professors in universities to determine with some accuracy the value of the instruction that they are giving.

So far we have been dealing only with the measure of efficiency of a single course; the measure of efficiency of a course in giving knowledge and command of the subject-matter covered by that course. We have not considered the educative value of that subject-matter, but merely the efficiency with which that subject-matter is given. We need more than that, and we need it very much. We need to know the efficiency of different subjects in developing general intelligence, and in preparing for the various pursuits in life, and that is a far more difficult problem, but one I believe not wholly insoluble. Here again we are in the habit of relying almost wholly on deductive reasoning. It stands to reason, we say, that a familiarity with natural science is the most valuable thing for a physician, a knowledge of history for a lawyer, of economics for a merchant, a skill in handicraft for an engineer, and may I add, of stenography and typewriting for a clerk, if not, indeed, for everybody in the world; and so on. Perhaps these assumptions are true; perhaps they are true in part. But in any case we ought to know by actual measurement whether they are true, and how far they are correct, and with the large number of young people that we are training, we ought to be able to measure results with an elimination of both accidental and systematic errors. We can do this by statistics, following the careers of those who have pursued different studies, with similar grades of proficiency, noting their progress in subsequent studies, professional or otherwise, and perhaps even in their chosen vocations in after life where other favoring or impeding factors do not come in.

If this were done on a really large scale by skilled statisticians, results of vast value might be obtained. As one of my predecessors many years ago said, "The object of statistics is to refute other statistics." It must be done by trained statisticians, or any one is led into courses of error, because they fail to eliminate the systematic errors. The accidental errors in large measure

cancel themselves out, but the systematic errors will show, and vitiate the result, if you are not careful to guard against them. If we take things on a scale large enough, and by a skilled statistician, the result will be of vast importance.

Take, for instance, boys out of school, who have now been going for many years on quite different courses. They are going to college; then they are going to leave college and the university. Something, of course, has been done in this line, but it might be done very much more, on a larger scale. Professor Dearborn did a great deal about schools and colleges in regard to rank. It might be done in regard to those boys who pursued classical, linguistic and commercial education, and their subsequent careers. Of course, you must avoid systematic error. We all know at the present day, if you took those figures crudely, that the classical man would appear best. You know right off that the classical man would on the whole show better in the college or in the professional school. You will find that they not only take better examinations in Greek, but they do better in other studies. Why? It is very simple. Because they have come from the best preparatory schools, and the best preparatory schools are teaching Greek. You have to eliminate errors of that kind. If you were to take that, you would find that the boy who took the classics in his youth, on the average went to a better school; on the average inherited more capacity; on the average had more impulse towards education; lived in a more educated home; had, in general, more advantages. You would have to guard against that.

To those who are interested in measurements of efficiency it is a source of great gratification that the Department of Economics at Harvard has requested the Department of Education to investigate the efficiency of its teaching. As yet, it is too early to speak of the methods that are being employed, for they are only under operation, and some of them not yet, but only projected, but it is fair to say that they promise much. At present, it is enough that by cordial co-operation a step has been taken which bids fair to bring education nearer to the goal of an exact science.

Measurements of Efficiency in Elementary

and Secondary Schools

BY FRANK E. SPAULDING, SUPERINTENDENT OF SCHOOLS, NEWTON, MASs.

T

K÷ODAY the most important and interesting aspect of the educational problem, both in discussion and in practice, presents a remarkable contradiction, -a contradiction that would be amazing were we not so thoroughly inured to contradictions in our profession. Everybody is clamoring for efficiency; almost everybody is claiming efficiency, and constantly increasing efficiency-for that bit of educational effort, whatever it may be, for which he feels himself to be immediately responsible. Yet no other proposition stirs so instant, so vigorous, and often so indignant protest, as the proposition to measure efficiency, a proposition which means the determination by objective measurement of the amount or degree of that thing which all are seeking, and nearly all are claiming, to the end that we may know whether or not we are really getting or producing what we universally desire, to what extent we are getting or producing it, and the progress that we are making in the direction in which we all want to move, and in which everyone of optimistic temperament thinks he is moving. It would be hard to conceive a proposition more simple, more practical, or more in keeping with the universal desire; indeed, it would seem that such a proposition must grow inevitably out of the desire and claim of efficiency.

Is it Possible to Measure Educational Efficiency?

The protest and attack which the simple proposal to measure efficiency is wont to provoke, are almost invariably and exclusively directed at the possibility of such measurement. Antagonists and protagonists of the proposal speedily find themselves engaged in heated discussion of the possibility of measuring the

efficiency of such things as personality and mother-love, a theoretical, metaphysical discussion, of the type for which our profession is notorious. There is need of no objective measure to determine the lack of efficiency, the utter futility, of such discussion. The profligate waste of thought, enthusiasm and energy, which such discussion involves, is at once ludicrous and pathetic.

We Are Measuring Educational Efficiency.

We need to come down to fundamental and pertinent facts. In the matter before us, the most fundamental and pertinent facts are these two. First, we are measuring efficiency all the time, we are measuring all kinds and aspects of efficiency, each one from his own standpoint. On the basis of the measurements of efficiency that we are all making constantly, we are determining our whole educational procedure, from the most comprehensive plan of organization,-institutional, municipal, state,―to the least detail of schoolroom method. We are basing a statewide system of secondary industrial education on our measurement and determination of the relatively low degree of industrial efficiency of the long-established, conventional high school, and on our measurement and determination of the relatively higher degree of industrial efficiency of the radically different subjects and methods of instruction that are brought into the industrial schools. We are adopting and discarding text-books on the basis of the measurements that we make of their efficiency in contributing to results that we desire to produce. We are measuring the degrees of efficiency with which instruction can be carried on in classes of different sizes and in various subjects; on the basis of these measurements, we are organizing classes of twenty-five, or perhaps only fifteen pupils, rather than classes of forty or fifty. We are measuring the efficiency of the face and general appearance-attractive or unattractive of the disposition and temperament, the sympathy, the enthusiasm, the ambition, the whole personality of teachers; and on the basis of these measurements we are engaging or rejecting, reëngaging or discharging, advancing or refusing to advance them in salary and position. That measurements that we are making are grossly crude, even that they are unjustifiable, inexcusable, intolerable, that they are largely or purely subjective and individual, we need