This paper outlines a methodology which can be employed to compare the empirical and theoretical hierarchies within a test instrument. The method is illustrated by its application within the profile and progression study carried out by the Assessment of Performance Unit in Science (APU) in the UK.
Some widely held hypotheses about performance hierarchies are empirically tested using the research data. These data were collected during the pretest phase of the study when pupils were at the end of their first year in secondary school (year 7, age 12).