The methodology of testing naive beliefs in the physics classroom

R D Donley,M H Ashcraft

DOI: https://doi.org/10.3758/bf03210922

Abstract:Many undergraduates harbor a variety of misbeliefs about physical objects in motion--for instance, that a bomb will fall straight down when dropped from a moving airplane. The evidence that these misbeliefs are resistant to correction by college-level physics courses, however, has often been based on methodologies that lack adequate internal validity. We used a quasi-experimental "before and after" design to assess the impact of two college physics courses, and we examined selection-bias, test-retest, and task-format factors directly. Initial accuracy and significant improvements due to instruction varied considerably by problem category and subject group; thus, in several ways, the results refute the general conclusion that conventional physics instruction does little to correct students' misbeliefs. We conclude by advocating the quasi-experimental approach for studies of naive beliefs in physics as well as for other situations in which the impact of classroom instruction is of interest.