Recently, our course team transformed a large-enrollment introductory genetics course from being predominantly lecture based to active learning based. During class sessions, students engaged in problem solving, which occurs when a student attempts to solve a problem without knowing the path to complete it. We designed class activities incorporating three distinct pedagogies from cognitive psychology: inquiry-based prediction, tell-then-practice case studies, and worked examples. We used a within-subjects design to compare students' attitudes toward these activities and their learning gains. Postsurvey results indicated that students felt worked examples helped them perform well on exams (a performance goal) and understand the information (a mastery goal) significantly better than the other activity types. However, students reported that all activity types required similar effort. Interestingly, students exhibited larger learning gains from prediction activities compared with worked examples or tell-then-practice activities, as evidenced by a course pretest/posttest. We discuss potential reasons for this misalignment between perceived helpfulness and actual learning gains. Additionally, we evaluate the strengths and weaknesses of each pedagogical approach.