ÖйúAƬ

So extra grades at A level are to separate the cr¨¨me de la cr¨¨me de la cr¨¨me from the cr¨¨me de la cr¨¨me. Does anyone really think that this can be done with any accuracy regarding future performance?

There is only one fair way to share a scarce commodity between too many who have an equal claim - by lottery. But we need evidence to support my contention.

I suggest that once the top A level has been graded into subgrades, a random half is chosen on the basis of the top of the top and the other half on the basis of a lottery of all with the previously undivided top.

Three years later, we will know which was the better selection process. If there should be no significant difference, then the more economical method should win. After all, life is a lottery and education is a preparation for life.

Lewis Elton
University College London