An Analysis of Group Interaction and Mathematical Errors in Heterogeneous Ability Groups

S ummary . This study investigated group interaction and mathematical errors of 11th‐grade students working individually and in four‐person heterogeneous ability groups. Averaging over all types of errors, high‐ability students did best in the individual condition, low‐ability students did best in the group condition, and medium‐ability students did equally well in both conditions. When types of errors were distinguished, for high‐ability students, the group condition was detrimental to learning new material (the mathematical algorithm) but was advantageous for performance on previously learned material (computational and algebraic manipulations); for medium‐ability students, the group condition was good for learning the algorithm but not for performance on computational and algebraic manipulations; for low‐ability students, group work was beneficial for performance of new and previously learned material. Group interaction was related to achievement: students who described or received explanations about the mathematical algorithm did well on the algorithm on the test, and students who explained how to perform computational or algebraic manipulations did well on those manipulations on the test.