Examines the extent to which grades derived from a conventional norm-referenced examination can be interpreted in terms of criterion-referenced performance assessments of different abilities and skills. Results suggest that performance is more affected by test format and subject matter than by the intellectual abilities tested by them. (JOW)