The Feasibility of Equating National Test Standards in Science between Key Stages 2 and 3 and from Year to Year

Abstract

The feasibility of equating marks on national tests in Science set at Key Stage 2 and Key Stage 3 is explored, using data from a final trial of the 1995 tests. Problems arise from curricular differences between key stages which lead to infringements of assumptions implicit in equating. Tests at different key stages are not strictly parallel and equatings valid for pupils from one key stage do not hold for the other. Moreover no single equating model seems likely to work in all contexts. However equating tests set in different years for a given key stage appears less problematic. The feasibility of equating is discussed in the wider context of setting standards for national tests and it is suggested that, for science at least, it is likely that comparability between key stages is likely to become of less concern than maintaining national test standards from year to year.