ABSTRACT
Before we look at ways of interpreting gender differences in pupils’ performance, let us first briefly consider bias – as a concept rather than a technical issue. There is consistency in the statements about bias in the national assessment documentation, i.e. that all efforts should be made to remove it. If assessors are to put this policy into practice, how are they to define bias? If we consider differences in group performance on an example of a reading test task (Hannon and McNally 1986), the authors report that over half of the children who had English as a second language did not select what was judged to be the correct response.
