Error Analysis as a Validation of Learning Progressions

Learning progressions (LPs) are a recent educational theory pertaining to student modeling. LPs argue that students with equal test scores may nonetheless have different conceptualizations of the material, with varying degrees of maturity. However, there is little empirical validation for LPs. To this end, we mapped two physics LPs (one predefined, one described in the paper) onto the answer choices of a popular conceptual physics test (the Force Concept Inventory; FCI). We then assessed 444 high school physics students using a pretest-posttest design. Students with more mature incorrect answers on the pretest performed better on the posttest than their less mature counterparts. We discuss implications for theorists and practitioners in learner modeling.