Abstract
In a paper in this journal, Backhouse (1976) described three methods of determining grades on a common scale for two groups offering a common paper in an examination. This is an important measurement problem and the need to find a solution to it has been intensified by current proposals for reforming examinations at 16+ and 18+, both of which envisage choice of papers according to level of achievement, taste, specialization etc. The purpose of this paper is to show that two of Backhouse's methods for dealing with the problem are unsatisfactory, while the other is likely to be unacceptable on non‐technical grounds. Worked examples to substantiate the objections are given. Finally another method for solving the problem is introduced.

This publication has 5 references indexed in Scilit: