I examined the impact of a self-diagnosis activity on students' conceptual understanding and achievements in physics. This activity requires students to self-diagnose their solutions to problems that they have solved on their own--namely, to identify and explain their errors--and self-score them--that is, assign scores to their solutions--aided by a rubric demonstrating how to solve each problem step by step. I also examined a common practice in the physics classroom in which teachers manage a whole class discussion during which they solve, together with their students, problems that students had solved on their own. Three 8th-grade classes studying force and motion with the same teacher participated. Students were first taught the unit in force and motion. Then a first summative exam was administered. Next, two classes (59 students) were assigned to the self-diagnosis activity and the other class to the whole class discussion (27 students). To assess students' learning with these activities, a repeat exam was administered. Results suggest that at least for teachers who are not competent in managing argumentative class discussions, the self-diagnosis activity is more effective than the whole class discussion in advancing students' conceptual understanding and achievements. I account for these results and suggest possible directions for future research.