Reducing Mathematical Misconceptions Through Cognitive Learning Principles

• In Event: Learning Sciences Poster Session

Abstract

We explore underlying mechanisms of incorrect worked examples, which have shown to be particularly effective for students with low prior mathematics knowledge. Students use a computerized tutor to complete a pretest, worked examples manipulation (designed to target fraction misconceptions), and posttest. Students are randomly assigned to one of five conditions that vary correctness of examples compared to a problem-solving control. Problem-solving practice with feedback was as effective as a combination of correct and incorrect examples. One example type alone was not as effective as a combination. Studying a combination of incorrect and correct examples reduced misconception strength which partially explained improvements on posttest. Further analyses reveal whether varying example types differentially evokes correct reflections which may also explain condition effects.

Authors

• Christina Areizaga Barbieri, University of Delaware

• Brianna Devlin, University of Delaware