An interesting article by Steven Schwartz on the currency of Grades and the problem of grade inflation at A level:-
“Sir Thomas Gresham, financial adviser to Elizabeth I, is remembered for his "law" of monetary economics - bad money drives out good. According to Gresham's law, if two kinds of money have the same denomination but a different value (gold coins v banknotes, for example) the paper money will drive the gold coins out of circulation because people prefer to pay their debts with the low-value paper currency and hoard the gold coins. The solution is to reform the currency so that there is only a single standard.Academic grades also appear to be subject to Gresham's law - bad grades drive out good. The process begins with grade inflation. Academic grades are intended to be an impartial, albeit imperfect, measure of how well a student understands a subject. Good grading is careful, accurate and allows universities and employers to make meaningful distinctions among students. Good grading also allows students to assess their progress.
Grade inflation occurs when marks go up without any corresponding increase in academic achievement. Because the top grade is fixed, inflation causes grades to bunch up at the upper end. This year, more than 22% of students received an A grade at A-level. These students represent a large range of performance. Those at the top end know a lot more about their subject than students who just squeezed into an A. Unfortunately, this information is lost when all students are given the same mark. As grades continue to inflate, the amount of information they convey decreases. “
Gresham’s law is only really applicable to commodity currency, but A Levels are more like a fiat currency – their value is derived from the fact that government “decrees” that they be acceptable to universities. So Gresham’s Law does not really apply to A levels – it isn’t bad grades aren’t driving out good grades but rather the government that is imposing them on the education sector.
Thus the problem with A level grades is essentially the same problem as exist with all fiat monies – there’s always a temptation to inflate. Now with fiat money, markets and rules can remove this temptation from central banks. In the education sector. similar rules (and markets) can be imposed on grades, but only if they exams are very well defined (which often means multiple choice). This is the route typically followed by the US system.
However A-Levels by their very nature, involve a less rigid marking system which permits grading that is more reflective of students abilities. The downside of the system however is that it’s virtually impossible set enforceable and time consistent rules in governing the marking. The solution is surely therefore much same as when a banking system looses credibility – replace it with a commodity currency. In relation to grades this means a system that marks by rank i.e. the top 10% get A’s. Commodity currencies, suffer the problem that exogenous events, such as gold discoveries, can cause inflation – and in the same way, commodity grades can induce grade inflation if one year a particularly stupid cohort of students takes the exam. While this is unfair to students in other year, in the UK it’s hard to see how the pool of ability can vary by much from year to year – which makes it the ideal commodity to back a grade system.
So if you’re going to choose a flexible exam system, fiat grades are unworkable and commodity grades would seem to be the only option.
Posted by wardx107 at October 22, 2004 02:04 PMThis is an excellent economic view of the grade inflation; I had never thought about that.
This grade inflation can also be applied to other exams, like the american GRE, a graduate education requirement. The mathematical part is supposed to convey information about the mathematical reasoning abilities of the candidates, but, what happens when more than 25% of the candidates are above the 90% of the grades?
What I see is that people are allowed to repeat the exam as many times as they want. Even if a record is kept of these different attempts, this only worsens the situation. In this case, I would suggest not to allow exam repetitions within a longer time span, perhaps a year or two, perhaps never.
The other problem is related to what you mentioned in your post today. There is also cheating in this exam, because people know now that the only worthy grade for the mathematical part of the GRE is 100%. There is an escandal about questions and answers of the GRE being published in asian webpages, which generated an usually good performance rate in these countries a few years ago; that's why they are back to pencil-based exams.
Another idea I would come up with is to (somehow)standarize results: grades should be concentrated around the mean, or perhaps the "passing grade", and should smoothly deviate from that. Any other distribution should raise concern.
Posted by: Leda Swan at November 18, 2004 11:11 AMYes I'd agree about GRE - seems so blunt its pointless.
disagree about using the mean - if you only give pass fail then how do you differentiate job canndidates when you've got 600 with "20 passes".
It's in job search that the costs of inflation are bourne by employer and employee alike (bbut not the educational establishment!)
Posted by: Giles at December 27, 2004 06:30 PM