Grades and Examinations
Those professors who think it not very important that a college spend much effort policing academic dishonesty among its undergraduates also tend to think little, or sometimes harshly, of the system of giving students grades, honor-point averages, and allied inducements to study. For one thing, if there were no grades there could be no academic dishonesty as we now know it. Our grading system appears to create a class of criminals out of thin air. The man in the street, who does not attend college, cannot commit academic dishonesty if he tries; it is only a system of our own creation that has infected a generation with this disease.
The communist theorists of the 19th Century argued much the same way about private property, which seemed to them an artificial creation of mankind entraining the crime of theft. Theft cannot exist without this convention. "To each according to his need," they would urge, and if need is self-defined (who else can define it, without inventing "ownership" all over again?) then theft becomes meaningless. Abolition of theft was not the only benefit communism was expected to bring to civilization, of course, but further debate is beyond the scope of this essay.
The analogy between economic theft and academic dishonesty is fairly close, actually, since most forms of dishonesty are a species of theft -- plagiarism, especially. And the objections that can be raised to gradeless classes as a cure for plagiarism are much the same as the objections that can be raised to communism as a cure for theft. In the case of theft, it must be observed that private property has its value in causing people to maintain material things and use them efficiently, not to mention produce them in the first place, so that its abolition might cost society more by inefficiency than theft does by intention. Recent experience in American colleges suggests that easier grading standards (and relaxed discipline in other matters affecting the educational process) may be reducing the amount of education students actually receive by an amount that exceeds the educational benefit predicted by those who instituted these relaxations. The easiest grading standard of all, which is no grading, is what would be needed to render cheating meaningless; but the total absence of grades would be quite undesirable.
These comments are not capable of proof in a paragraph or two, and in fact it is denied by some thoughtful students of higher education (e.g. Milton et al, 1986) that the giving of grades has any value at all, at least when they are given by the semester, by the course and by the test, as our colleges do now. A defense of having a grading scheme of at least some sort is offered in [Raimi, 1967], where I have also put forward a description of a method of giving examinations and degrees that satisfies the legitimate purposes of a grading system while at the same time reducing essentially to zero the probability that the system can be used fraudulently (as our present system is used by student cheaters and plagiarists). What follows is a summary of that scheme and some commentary on what I have learned since publishing it more that twenty years ago.
The idea, in brief, it to divorce teaching from examining and grading entirely, and to give degrees based on a very few examinations. My proposed mechanism was not original, but was modeled on the system that was in use at the Sorbonne (Paris) when I was there in 1950. A student (in the system of [Raimi, 1967]) declares himself to be a major in some subject, say mathematics, and thereby subjects himself, to earn the Bachelor's Degree, only to the requirement that he amass a certain number of certificates, each one representing knowledge of a rather large part of his major field, except that two or three (of a total of, say, twelve certificates) will be in either an allied field or in some approved elective subject, according to the printed regulations of his major department. The certificates will have names, like Differential and Integral Calculus, Archetypes in Literature and Film, Rational Mechanics, Western Civilization 1492-1914, or Economics From Smith to Schumpeter, and will normally be awarded to students who pass the corresponding certificate examinations. And it would be expected that the average student could accomplish this in about four years.
The certificate examinations (the proposal goes on to suggest) are given recurrently, perhaps every spring, some of them in December as well (or instead), during examination weeks at the ends of the terms, and students take them when they think they are ready. There is no penalty, except lost time and a lost examination fee, for failure. If a student fails, he may try again another year. Even if he passes, he may repeat an examination for an improved grade. If a student doesn't pass his contractual list of examinations in some fixed period of time (maybe five years, with extensions by petition) he is dropped from the rolls. When he has passed the list he is given a degree, perhaps with honors or distinction.
What the student does from day to day is this: He attends lectures and discussion and problem classes and laboratories if they seem relevant to the examinations he will be responsible for, or if he enjoys them. His enrollment in the college permits him to attend anything he likes, subject to the inevitable priorities imposed by considerations of space and the autonomy of professors. While my initial proposal did not include this, criticism soon caused me to add a tutorial feature: If the college can afford it, and to the degree possible, students will be assigned to "tutors" or "advisors," who will do what their titles suggest, to help them in any way to pass the examinations they are preparing for. In particular, tutors will assign papers, problem sets, laboratory exercises and the like, and read the results and comment on them. They may even grade them, to give the student an idea of how he is doing, but these grades will have nothing whatever to do with the degree, or the transcript of record.
Naturally these same professors who serve as tutors, and as lecturers and discussion-leaders (even if there is no "tutorial" system), and who as such are allies of their charges, will be largely the same as those who set and grade the yearly (sometimes semi-yearly) certificate examinations, but not in one-to-one correspondence. In addition, I would hope for the help, in setting and grading both, of experts from outside the college, there being nothing that maintains standards better than the scrutiny of disinterested professional colleagues. Only rarely will a teacher become the judge, via an examination of his own devising, of the very student he has taught as tutor, and even then the examinations will have been composed and graded by committees. The examinations will be long and the answers discursive, and not such as can lend themselves to copying. They might be partly oral, or partly practical (as in a laboratory, or in music). The essays or exercises during term, set by the tutors or suggested by lecturers in courses, i.e. the things that correspond to what we now call "term papers" and "take-home examinations," will be only intellectually related to the examinations that produce the certificates. Their value as practice and their uselessness as documents would be so obvious that plagiarism could be only an exercise in penmanship or self-deception.
It was only in part the problem of academic dishonesty among undergraduates that led me to think about such a reform in the way colleges teach and evaluate students; there are other educational and administrative advantages to the scheme. Here is what I said to the University of Rochester Faculty Senate on January 9, 1967:
It would cause the student's attitude to his own education to become more comprehensive, and less regimented. At present he has to pass each four courses on schedule. He has little freedom each day to penetrate anything deeply, and he is encouraged to give all his attention to the multitude of small hurdles he must cross daily or weekly. This induces the attitude that small hurdles are what make up his life.
The proposed scheme would direct his studies towards larger exhibitions of competence. He would have some reason to read in the summers and vacations. He would be able to worry a point until he got it, rather than have to put it aside because of other pressures, and hope that a superficial outguessing of the professor's next exam will make it unnecessary to do more.
It would relieve some of the 'exam pressure' students now feel. There would be a second chance for all, which is under the present system technically impossible. The second chance is not only for failure, it would give an accurate rating to students who grow while they are here. A senior history major repeating a freshman-level history exam could -- or should -- get an A unless he is a dolt. He should retain the hope of that A however he began his career, and he deserves to be reported out to the world as having rated that A if indeed he is, at the time, that competent. Our exams are not intended, after all, to be intelligence tests. But they often are in effect irrevocable reflections of temporary incompetence.
It would separate the function of the teacher from that of the examiner, permitting the student to look on his teacher as an ally and not an examination-day enemy.
It would practically eliminate cheating.
It would enormously simplify record-keeping. [Here I gave some examples, involving transfer-students, advanced placement, night school, Junior Year Abroad, illnesses, "dropping" courses, etc.
It would clarify, in each department, what they think they're teaching, and prevent many duplications.
Today I can add an advantage I did not have the experience to recognize in 1967: The certificate scheme would permit the student to learn from professors whose lectures are genuinely helpful, and to avoid the others. A professor whose lectures or labs are known to be unhelpful will discover this from his attendance figures, and he may be induced to improve, without being called on the carpet by a Chairman or Dean. Today such a professor is protected from such knowledge by the fact that his is a "required" course, or by his habit of giving high grades to everyone.
This régime of giving degrees by examination does not preclude the giving of 'non-credit' examinations in individual courses, if the professor thinks these are pedagogically wise; indeed, the students are likely to demand such examinations, since they want some advance warning of how the certificate examinations will work. Students will also demand writing exercises, reading lists, laboratory manuals; and as mentioned earlier, one would hope for a substantial number of individual tutorial sessions, with associated writing assignments.
All this seemed quite possible to me when I first wrote it out in 1966. My Fulbright year at the Sorbonne had not in fact been spent this way, since I was not interested in a Sorbonne degree, but my friend Walter Nöll, now a distinguished mathematician, did manage to pass all the seven required examinations for the Licence (a sort of bachelor's degree) during his one year there. He had of course already been studying mathematics (in Berlin) before that year; nobody could learn that much that fast. What I myself learned that year is very hard to say; probably very little, though I attended a lot of lectures. Without examinations, I could not myself be sure.
French students seemed to work well enough by this system if they were interested to do so. Some became academic bums, of course, "perpetual students" as in a novel by Turgenev, but that was their look-out. There was more of this in Paris than in (say) Oxford, because of, in Paris, the lack of selectivity in admissions, the large classes, and the lack of individual tutorials. On average, though, students in France seemed somehow to be more adult than American undergraduates, and I attributed this to their self-paced educational system at the University.
In fact, before the twentieth century this apparently Utopian scheme was pretty much the way most universities worked; only recently have college courses come to resemble high school classes, with a multitude of grades given by a multitude of individual professors, based on an even larger multitude of graded small assignments. Even today, at Oxford and Cambridge, enough of the older attitude prevails so that the examination system, while somewhat different from that of the Sorbonne in 1950, is still much more like what I was proposing than anything to be found in the United States. It worked in 1962, when, spending a Sabbatical year in Cambridge, I last saw it in action, and it seemed to me that their régime of long springtime examinations set and scored by committees directed students' attention much more towards what really counts in education than does the American system I was brought up on, and was now myself as a professor applying in Rochester.
So I presented my paper to the Faculty Senate and asked for a debate. A few professors were enthusiastic, and ready to reconstitute the whole curriculum on the spot, but the great majority saw insuperable difficulties.
A professor of music disagreed about the union of many small things (as now taught and tested) not adding up to anything valuable, and explained that in his discipline (music education) "much of the knowledge is necessarily broken down into smaller segments and mastery of major skills into mastery of sub-skills."
A professor of English (I quote from the Minutes of the Senate meeting) "thought that the premium placed on the examination, especially as it could be taken more than once, would remove the urgent incentives on students to attend classes, to fulfill written assignments, and to study systematically...and he was afraid that the only gesture in the direction of things academic the majority of students would perform would be participation in one gigantic cram session on the eve of the May examinations." He did think that associating the examination system with a severe tutorial system, as at Cambridge, might work here too.
A professor of engineering agreed with the professor of English, saying that while the aims of the proposal were exemplary (everyone seemed to agree on that), it "places a premium on the maturity and motivation of the students...it was naive to expect that any but a minority of the students would by themselves develop proper study or work habits.”
A professor of education considered the proposal to give the better (or most recent) of the grades to a student who took a certificate examination more than once to be a distortion of the record; he thought the University should consider carefully before deciding between announcing typical performance, best performance, or most recent performance, of its graduates. He also failed to see that cheating would be diminished. With only twelve chances to cheat, a clever cheater, he thought, could amass more fraudulent points, as a fraction of his college career, than under the present system.
A professor of psychology considered the proposal to be a "set of intuitive responses, resulting in a set of speculations about how the educational process might be improved" and suggested that there was no use discussing them until prior questions such as "How do we learn?" have been answered.
Another professor of engineering expressed doubt that the "system could provide enough statistical data for proper evaluation of the students." He had, from experience of his own, which he described, reason to believe that there might be little correlation between the results on these comprehensive (certificate) examinations and the performance of those same students in "course work."
So it went. Some professors offered compromises. One of them amounted to asking for graded course work of a certain quality as a prerequisite for entrance to the certificate examinations. Another was to use the scheme "in the last two years," i.e. the time a student is "in his major," with the introduction to higher education, i.e. the "first two years," paced and disciplined in the present manner. Another thought it a good scheme at the Master's degree level.
My paper [Raimi, 1967] was published in the Bulletin of the American Association of University Professors and attracted a good bit of mail, two invitations to speak at other colleges, and an invitation to appear before the Commission on Liberal Learning of the Association of American Colleges in Washington, D.C. So far as I know, nothing whatever has been done by any college whatever (including mine) in the direction of my proposals. Indeed, the movement for the next fifteen years was all in the opposite direction.
Rochester in 1967 still had on its books the vestiges of a system of "comprehensive examinations" in the senior spring semester. Each Department was supposed to cause its majors to study for these examinations, which could be oral or written or both, exactly to accomplish one of the things my certificate proposal intended: to induce in students an overall view of what they had learned, and to be sure not to graduate anyone who had not achieved it. By the time I arrived at Rochester, in 1952, the "comprehensive exam" system was already on the rocks, at least in mathematics. Nobody ever failed, and nobody I saw in these examinations had got anything out of the supposed exercise. By 1967 the University had permitted most departments to replace this time-wasting by an "approved substitute," which in mathematics was participation in a special senior seminar. That wasn't bad, but it too soon evaporated, as did so much else, in the student revolutions of the next seven years.
From 1967 to about 1982 there took place a national grade inflation, and students in general became subjected to ever fewer limitations on their autonomy in curricular choices, e.g. in the way of language and "distribution" requirements. Unified standard first-year requirements like "Western Civilization" tended to fade away. Most recently there has been a slight movement in the other direction, especially in demanding of everyone more training in writing than before, but the newer requirements tend to be in the direction of diversity rather than insistence on a common core of knowledge. That is, faculties now want to make sure that everyone has a certain little bit of "science" and another bit of "humanities," and not be too narrow; also that if Civilization or Literature is to be required, it should contain non-Western elements, and other broadening experiences, like Black and women's literature and Oriental history and cultures.
New departments and programs have arisen, most of them called "interdisciplinary," such as women's studies, black studies, film studies, computer science, cognitive science, policy studies, religion, and regional studies of several sorts. Thus "religion" as a program requires, in addition to the major study of the religious teachings themselves, some allied work in philosophy, history, certain ancient languages, and possibly anthropology and sociology, according to a formula worked out by the program directors. These programs make a certificate examination scheme increasingly impossible, even if we could assume the maturity that would keep undergraduates from wasting their time, and even if we could afford a thoroughgoing tutorial system.
There may have been a time when it was clear what a degree in economics, or physics, or French literature might mean, and one could imagine the faculties of the respective departments agreeing on what the certificate examinations should ask, and what the answers should look like. Today we have degrees given for studies of such enormous variety, what with internships, Junior Years here or there, interdepartmental projects and things called practicum, that there is a large class of professors with an interest in preserving the credit-value of their special offerings and fearful of subjecting the beneficiaries of their unique insights to a unified and possibly unsympathetic scrutiny via a certificate examination. And their students feel the same way.
I have described my proposal to undergraduates from time to time. Students of physics or economics often find it appealing, but a student of psycholinguistics with four hours' credit in each of various elementary courses in philosophy, Italian, human sexuality, statistics, feminist poetry and social problems, among others, cannot imagine how his education would be judged under that scheme, or how he would use these things in preparing for examinations of the scope described, whatever their titles. I can, however; and in fact I believe that any worthwhile course of lectures, or worthwhile exercise in writing small research paper, would be of value in some later examination. But that would take an effort of intellectual integration not demanded of our undergraduates today, and probably beyond our asking.