Performance Expectancies Moderate the Effectiveness of More or Less Generative Activities Over Time

Area of Science:

• Cognitive Psychology
• Educational Psychology
• Learning Sciences

Background:

• Learning effectiveness can depend on individual differences, such as performance expectancies (PEs).
• Generative learning activities, like problem-solving, are often contrasted with less generative activities, such as worked-examples.
• The long-term benefits of generative learning may be moderated by pre-existing student beliefs about their abilities.

Purpose of the Study:

• To investigate whether the long-term learning benefits of generative activities depend on individual differences in performance expectancies (PEs).
• To compare the effectiveness of greater generative activity (problem-solving) versus less generative activity (worked-examples) on immediate and delayed student performance.
• To examine how different learning activities influence changes in students' performance expectancies and metacognitive accuracy.

Main Methods:

• Eighth-grade students (advanced track) were assigned to either a problem-solving or a worked-examples condition.
• Immediate and delayed (3-month) performance tests were administered on linear functions.
• Individual differences in performance expectancies (PEs) were measured prior to the learning intervention.

Main Results:

• A three-way interaction between PEs, generative activity, and time was observed.
• Initially, problem-solving benefited students with higher PEs, while no difference was found for lower PEs.
• In the delayed test, problem-solving outperformed worked-examples for students with lower PEs, but not for higher PEs, whose performance declined over time.

Conclusions:

• The effectiveness of generative learning activities is not uniform and depends on the interplay between the activity type, student expectancies, and time.
• Problem-solving may lead to more accurate self-assessments and reduced overconfidence, particularly for students with initially lower performance expectancies.
• Learners with higher PEs may experience an initial advantage with problem-solving, but this benefit may not be sustained long-term compared to worked-examples.