How did/does the Greek-letter grading system (formerly?) used at Oxford and Cambridge work?

Question

Does anyone have any information about the alpha-beta-gamma marking system? What are/were the criteria for the different grades and how were these then collated into degree-class marks? Is this system still in use?

Addition:

Thank you for all these answers. I unfortunately neglected to specify that I meant something other than the use of alpha and beta quality marks in maths at Cambridge.

Based on one paper I've read (Stray, "The Shift from Oral to Written Examination: Cambridge and Oxford 1700–1900"), fleeting (parodic???) mentions of marks like "beta plus plus query plus" and "alpha gamma" (I seem to remember Parfit describes Kant's ethics like this somewhere…), and personal reminiscence (although I think I might remember one MML supervisor at Cambridge in the 90s marking like this, I definitely do remember conversations with people from Oxford where they spoke of needing a certain number of alphas to get a first—akin to what is described here https://users.ox.ac.uk/~manc0049/guide/guide_6.html), I think it was more of an Oxford thing.

Answer 1

This is anecdotal, about Oxford in the period 1970-2000.

Being a mathematician I did not use this code: we expressed our examination marks in the form N alpha^m beta^n, meaning an overall points total of N, with m questions almost complete, and n questions with substantial answers. However, I lived among those who did use the code, and think I understood it in broad outline.

Basically the code was developed in the first half of the 20th century. At that time students in Oxford were classified in their final examinations as gaining Honours in the First, Second, or Third Classes. (There were also special cases: a Fourth Class; one might not get Honours but be awarded a Pass and so get one's degree; or one might Fail.) Basically alpha means First Class, beta Second Class, gamma Third class). But although the final consolidated mark was I, II, or III academics cannot resist making more and more refined distictions and so a more elaborate system developed.

alpha

alpha -

alpha --

alpha ---

alpha beta

beta alpha

beta +++

beta ++

beta +

beta

beta -

beta --

beta ---

beta gamma

gamma beta

gamma +++

gamma ++

gamma +

gamma

gamma -

gamma --

gamma ---

A further elaboration was that as a further refinement the final + or - might be preceded by a ? sign: so for example beta +?+ is better than beta + but not as good as beta ++.

I think that sometimes alpha + etc were used, but purists insisted that it was impossible to be more excellent than alpha.

I do not think that delta was used seriously: as I recall it "s." [ i.e. satis] denoted "Pass" and "n.s." denoted failure.

I think that a mark alpha gamma really signified that the examiner could not decide whether this was brilliance or nonsense: such a mark would have to be resolved to something on the scale before being used.

You ask for the criteria for achieving these marks. That was not how academics at that time operated. Each student was somewhere in one of the Classes: the examiner just had to discern where, and allot the correct symbol. But do not think that this was a random process. When in the late 20C serious studies were made to discover how consistent the assessment was in given subjects, it became clear that there was a very high degree of agreement between different markers.

In Finals a student would do usually eight 3-hour written papers, answering in each perhaps four questions from a menu of perhaps a couple of dozen. Each paper would be read and independently given a mark by two examiners; if there was a discrepancy the two examiners would discuss and produce a reconciled mark.

The whole panel of Examiners (six or eight) of a subject would then meet and produce the Class List. By the eighties there were usually overt conventions about how they would use the marks of the eight papers to do this. But there were a variety of conventions. However they usually looked at the overall (dare I say average?) picture and then required for a First a minimum number of leading alphas and a maximum number of low scoring papers; and similarly for the other classes.

A disconnect with reality occurred when the University divided the Second Class into II-i and II-ii, and the decline in the number of III awarded, and subsequently the decline in the number of II-ii. Although the system may still be used informally all official processing of examination marks in now (I believe) done in terms of University Standardised Marks. This is a 100 point scale where, to comply with national protocols, 70 signifies a I, 60 a II-i, 50 a II-ii, 40 a III, 30 a Pass.

Answer 2

I am only aware of this being used for maths at Cambridge (and I think is still in use, although the specifics have probably changed over the years).

It is to encourage answering a small number of complete questions rather than trying to pick up a few marks on many questions. This is necessary because of the peculiarity of the Cambridge exam system. Unlike other universities, where you choose what modules you study and then only sit exams in those modules, at Cambridge all students sit the same exams, which have questions available on all modules, and you just answer the questions relevant to modules you have studied. So there are many more questions available than any student would be able to solve in the time available. Normally much more work and understanding is required to get the last half the marks on any question than the first half, and so without this system the best strategy might be to get a basic understanding of many courses, enough to do parts of questions but not necessarily enough to actually answer a full question.

The system was something like, for a question marked out of 20, 15+ marks would get an alpha, and 7+ would get a beta. These would be bonuses, in addition to your marks on the question, rather than replacing them. So your final score would be X number of marks, Y alphas and Z betas. For some shorter questions you would only be able to get a beta. This information was, of course, widely available to students.

For the division between first class and upper second, the most important thing is the number of alphas. For lower divisions, or for borderline cases, number of betas comes into play, and even further down is total number of marks. The boundaries have sometimes been based on specific linear combinations of X, Y and Z.

The taught masters (the fourth year, hence known as "Part III") has a more normal modular system. However, modules still get graded as alpha, beta or gamma (gamma being positive marks but less than beta). In fact they could also be modified here (but not in the undergraduate case) - you can get alpha-minus or beta-plus. Gamma is not used in the undergraduate course.

Answer 3

I did maths at Cambridge in the mid 90s. Back then at least

• Alphas and betas (I don't remember gammas) were awarded for how thoroughly & correctly you answered individual questions. In addition to the normal mark-scheme points: if you completely answered the question you got an alpha, and a near-miss was a beta.
• To get a particular class you needed a number of alphas in addition to enough points. If you had a mismatch then your paper went to moderation.

The idea being (I assume) to require deep knowledge in a few areas and not just breadth. I got the impression this was maths only though, and other subjects didn't use it.