Saturday 19th January 2013

This is an old chestnut that I keep on getting asked - notably by students wanting to know how many marks they need to get a given grade and by parents who are baffled by the constant references to 'grade inflation' in the newspapers.

To answer this question it is probably easiest to look at how exams. are NOT marked nowadays.

Back in the day when I (and many older parents) took O-levels and A-levels the system for grading papers was known as 'normative marking'. It is still used by a lot of countries apart from the UK, is totally transparent and is a pretty easy concept to grasp even for a mathematical duffer like me.

Under 'normative' marking all exam scripts are marked according to a pre-established set of mark schemes and a raw mark is given to everybody's paper. Then all of the raw marks for everybody taking that examination are put together and all candidates who get marks that are in the range of marks achieved by, say, the top ten percent of the cohort taking the examination get a grade 'A' - irrespective of what those raw marks are. Everybody whose raw marks are, say, in the next ten percent of the marks achieved by the whole cohort get a grade 'B' and so forth.

This system meant that candidates were marked with reference to the achievement of everybody taking the examination and that everybody knew that if you got a grade 'A' you were in the top ten percent of the country in terms of raw marks achieved in that sitting. Of course, the raw marks needed to be in the top ten percent (and get a grade A) varied from year to year as the ability of the cohort varied and as the 'difficulty' of the examination varied.

The system had its pros and cons. For employers and universities looking for 'the best' it meant that they could be sure that a grade 'A' candidate was in the top ten percent of the country. However, given that grade 'C' was set as a 'pass' at O-level and a grade 'E' as a pass at A-level it meant that 70% of all candidates would automatically 'fail' a given O-level and 50% of all candidates would 'fail' a given A-level. It also meant that the raw marks required to get a given grade would often be very tightly bunched around a given figure and that the difference between a grade 'A' and a grade 'C' could be remarkably few raw marks.

In an attempt at mitigating these issues the actual percentages used to band British A-levels into grades did not go down in ten percent increments - the figures actually used were: A, top 10%; B next 15%; C next 10%; D next 15%; E next 20%; Fail last 30%. A-levels used to have the 'consolation prize of an 'O' grade (which stood for 'O-level pass') for the penultimate 20%, which in practice meant that only the lowest 10% of marks failed A-level completely. I cannot comment on the actual old O-level percentage grade boundaries but I imagine that they were similar.

Enter 'grade reference marking' (at about the same time as the introduction of GCSEs in 1988). The idea behind grade reference marking was that instead of candidates being required to obtain marks that lie within the top ten percent of marks of all of the candidates to obtain a grade 'A', the candidate simply needed to produce a 'grade A standard answer'. It is at the (possibly rather subjective) whim of the examiners to decide what constitutes a 'grade A standard answer'.

This system (the one currently in use in the UK) also has its pros and cons. It does mean that in theory everybody taking that examination, if they are good enough, can get a grade 'A'. This is extremely good for the morale of candidates who might have crashed out completely for the lack of just a few raw marks under the old system, but it does make the task of employers and universities a great deal more difficult. In short, if almost everybody is getting a grade 'A' then how are employers and universities to know which candidates are the 'best'?

It is no great surprise that the rise of grade reference marking has led to the re-introduction of university admission tests and various other non-public-examination-based tools to select the best candidates for employment or higher education. However, this fact does rather beg the question as to whether public examinations are still fit for the purpose for which they were designed. A great many people (including myself) would argue that they are not and that we need to return to some sort of normative marking, preferably administered by a single national examination board rather than the fragmented British 'system' we have today.

Another, more sinister, aspect of grade reference marking is that it is not at all a transparent system and it is totally at the arcane whim of the examiners and their statistical advisors to decide what constitutes an 'A-grade standard answer'. It is facile nowadays to say that the examination boards can be trusted to administer the system fairly and honestly as they have all been caught with their hands in the cookie jars at various times 'adjusting' various grade boundaries when it has suited their purposes to do so (more on this in later blogs). Students and parents seem to have difficulty imagining that such trusted 'establishment' institutions as examination boards can lie and cheat, but with public trust crumbling in politicians, bankers, policemen and other once-respected members of the British 'establishment' perhaps reality will soon intrude?

In future blogs I shall look at the various ways in which examination boards have cheated in the past and the things that parents, teachers and students need to be on guard about. I shall also look at the confusing 'UMS' (Uniform Mark Scheme) system and how it has been manipulated, and present a few figures regarding marks-to-grades conversions for various subjects that are quite astounding. Kafka could not have invented a lot of this!

The British are a remarkably docile lot, not given to storming the barracades. Although these are scandals of national dimensions affecting the educational progress of tha nation's most precious resource - its young people - it is depressing how few returns Google searches on, for example, 'Normative Marking' will bring. Most students and parents simply do not care to try to understand what is made a deliberately very complicated and secretive system. I find that extraordinarily sad. Am I a lone voice crying in the wilderness? Maybe, but I feel that it's better than simply sitting back and taking this national scandal in total silence.