An earlier study of gendered achievement at the UWI School of Medicine at St. Augustine suggested that most differentials were negligible to small at the preclinical stage and small to moderate at the clinical stage (De Lisle and Pit-Miller 2002). One explanation for these small gender differences was that the stringent selection processes created sample restriction. In 2003, changes were made to the assessment scheme in the biomedical sciences, with the elimination of a number of formats.

This paper pressents guidelines for conducting a gender analysis within the formal educational system. The need to reach beyond indicators which can be readily obtained from population census data and to explore, in depth, the actual process of education, and the differential consequences of this process for female and male students is emphasised.