Some measures of academic attainment at GCE (Advanced) level

Publisher Website
Google Scholar
Abstract
Any numerical parameter relating to something as complex as ‘academic ability’ must be regarded with some reservation. Although statistical methods may sometimes improve the interpretation of numbers, they can never completely eliminate the fallibility of the primary data and may even obscure this basic weakness. In this paper we have made a number of assumptions, such as that aggregate A‐level scores reflect academic ability, and that if aggregate ‘O’ scores show a good correlation with ‘A’ scores, then these, too, must reflect the same. These are broad generalizations, and any experienced teacher will be aware of their limitations. For instance, O‐level scores are affected by many other factors such as schools, teachers, groups, social background, maturity, etc. In addition, one may ask if an ‘A’ grade in O‐level Domestic Science or Drama is the same indicator of academic ability as an ‘A’ in Physics or French. With these reservations in mind, we have developed a scheme for comparing A‐level performance between different subjects in this college by taking account of two contributing factors, namely O‐level scores and the ‘degree of severity’ of the A‐level subject, as revealed by match‐pairs statistics. Our analysis has been both interesting and informative, in some cases confirming what we had long suspected, whilst in others throwing new light on examination results. The tables have also provided a useful focus for discussion between the principal and heads of departments.