The methodology developed by Anderson (1982) within the framework of the Information Integration Theory is applied to the study of 13‐14 year old students' mastery of an elementary concept in the high school physics curriculum, Archimedes' Effect. Of the five factors of variation tested in the experiment, (a) the volumic mass of the liquid, (b) the volume of the immersed object, (c) the volumic mass of the solid, (d) the horizontal or vertical position of the solid and (e) the depth of immersion, only a and c were systematically taken into account by a majority of the subjects, whose task was to judge the strength of the effect in different situations. Individual differences were pronounced. Six groups could be formed on the basis of the importance assigned to one or two factors. The use of the Anderson paradigm in learning or classroom testing situations is discussed in the light of the results of this study.