Meta-analysis of problem-based learning

Researchers from The Netherlands recently re-examined studies of problem-based learning with post-secondary students. Their aims were to study the main effect of problem-based learning on the acquisition of knowledge and skills and to determine which factors influenced learning in this environment.

In contrast to the conventional teacher-led, lecture-based instruction, problem-based learning consists of student-centered, self-directed learning in small groups, with a teacher or tutor serving as a facilitator or guide. Working with authentic problems, students acquire the knowledge and skills necessary to solve similar problems in the real world. Because analyzing and solving representative problems is at the core of problem-based learning-life problems. The Dutch researchers questioned whether the knowledge and skills acquired in problem-based learning were greater or more effective than the learning achieved through conventional instruction.

Forty-three studies met the statistical criteria for inclusion in this analysis. Three-quarters of these studies measured the knowledge students gained and one quarter measured the skills students acquired to apply their knowledge. Results suggest that students in classes using problem-based learning acquire less factual knowledge than students in traditional classes.

In problem-based learning, students are required to immediately apply their knowledge to problems. This is not true in conventional classrooms; only after two years of study in a conventional curriculum was the application of knowledge emphasized. Thus, the difference in the knowledge level between the two groups of students was greatest in the first two years. After two years, however, students in problem-based classes began to catch up to the conventional students’ level of knowledge. In terms of application of knowledge or skills, the students in the problem-based curriculum significantly outperformed conventionally trained students at all levels.

In addition, students trained in problem-based learning retained their knowledge better than conventionally trained students. In conclusion, although students in the problem-based learning programs have slightly less factual knowledge than students in conventional classes, their knowledge is more detailed and they remember it better over a longer period of time.

The two programs assessed the knowledge and skills of college-age students in very different ways. These researchers believed that the type of test used to measure achievement can produce conflicting results. Therefore, researchers analyzed to the degree to which tests in these studies assessed students’ knowledge of facts or evaluated their skills in applying their knowledge. Results revealed that the better an instrument is at evaluating students’ skill in applying knowledge, the more positive the results for problem-based learning compared with conventional learning environments.

Significant positive effect from problem-based learning

These researchers conclude that there is a significant positive effect from problem-based learning on the skills of students. No study reported negative effects on such skills. However, there is a tendency for negative results with problem-based learning when evaluating the factual knowledge of students. Two studies out of the 43 were mostly responsible for this negative finding. Further analysis revealed that the expertise level of students is a factor in the size of this difference. In the early years of study in a subject, the conventional curriculum produces greater learning of facts, but this difference decreases as students take more courses. In summary, students in problem-based learning classes have slightly less knowledge, but remember more of that acquired knowledge than students in conventional classes. Students in the problem-based learning classes show a better ability to apply their knowledge at all levels. The type of assessment is key in determining results. Clinical performance and skills of students in problem-based learning were superior to those of students educated in a traditional curriculum. Tests of basic science factual knowledge tend to favor traditional teaching approaches, at least in the earlier years of training.

“Effects of Problem-Based Learning: A Meta-Analysis”, Learning and Instruction, Volume 13, Number 5, October 2003, pp. 533-568.

Published in ERN October 2003 Volume 16 Number 7

Leave a Reply

  • (will not be published)