When is Problem-based Learning More Effective? A Meta- analysis

Problem-based Learning as an active and constructive teaching method in the educational system, drew the attention of researchers in recent decades and many studies have been conducted to evaluate its effectiveness that have reported mixed results. Therefore, this meta-analysis was performed in order to achieve more accurate judgement about the effectiveness of PBL. The study population included all studies that has been conducted since 1384 which measured the effect of PBL on student performance that are available in databases.  The evaluation is based on inclusion and exclusion criteria, 27 studies were selected and were analyzed using the software CMA2. By performing the sensitivity analysis in order to remove biased published studies, 73 effect sizes remained. The combination effect sizes of fixed and random models are 774/0 and 723/0. This means that PBL is moderately effective. Based on the analysis of moderator variables revealed that PBL is more effective on skills, in university classroom and compared to traditional methods.