Method to trace coherence and persistence of preconceptions

The development of preconceptions throughout secondary education is studied by analysing performances on tests administered in forms 1 to 5 of a school in The Netherlands. Scales of questions on mechanical interactions and on electric circuits are constructed, by analysing answering patterns of pupils choosing distractors alternative to the physically correct answers. Distractors which are plausible on the basis of the same idea or line of reasoning are grouped into a particular alternative scale if they load on the same factor. The alternative scales traced in the mechanics test refer to preconceptions associating force with objects ‘inducing activity’ or ‘having dominant properties'; the preconceptions on electric circuits concern ‘current consumption’, ‘constant current supply’ and ‘local and sequential reasoning’. Throughout consecutive years of schooling both the physical and the alternative scales appear to increase in reliability; the various preconceptions however, as indicated by regression analysis, differ significantly in their degree of persistence.