BACKGROUND: In problem-based learning (PBL), high-quality discussions are crucial for student learning. The quality of the discussion is affected by the quality of the contributions students make during PBL tutorials. AIM: This study investigated whether students are able to evaluate the activities of their peers in PBL groups in a reliable and valid way. METHOD: For this purpose, the Maastricht-Peer Activity Rating Scale (M-PARS) was developed. The M-PARS was well-founded on the literature about effective tutorial performance. With this scale, students (N = 196) were evaluated by their peers on three main aspects: their constructive, collaborative, and motivational contributions to the group. RESULTS: A confirmatory factor analysis demonstrated that the data fitted the three-factor model reasonably well. The generalizability studies demonstrated good internal consistency when students were evaluated by, at least, four of their peers. Furthermore, Hancock's coefficients indicated good construct reliability. CONCLUSIONS: The results prove that peers are able to provide reliable and valid information about a student's active participation in the tutorial group, if at least four peer ratings are attainable, out of a group of eight students. In conclusion, the M-PARS is a valid and reliable instrument.
BACKGROUND: In problem-based learning (PBL), high-quality discussions are crucial for student learning. The quality of the discussion is affected by the quality of the contributions students make during PBL tutorials. AIM: This study investigated whether students are able to evaluate the activities of their peers in PBL groups in a reliable and valid way. METHOD: For this purpose, the Maastricht-Peer Activity Rating Scale (M-PARS) was developed. The M-PARS was well-founded on the literature about effective tutorial performance. With this scale, students (N = 196) were evaluated by their peers on three main aspects: their constructive, collaborative, and motivational contributions to the group. RESULTS: A confirmatory factor analysis demonstrated that the data fitted the three-factor model reasonably well. The generalizability studies demonstrated good internal consistency when students were evaluated by, at least, four of their peers. Furthermore, Hancock's coefficients indicated good construct reliability. CONCLUSIONS: The results prove that peers are able to provide reliable and valid information about a student's active participation in the tutorial group, if at least four peer ratings are attainable, out of a group of eight students. In conclusion, the M-PARS is a valid and reliable instrument.
Authors: Rachelle J A Kamp; Diana H J M Dolmans; Henk J M Van Berkel; Henk G Schmidt Journal: Adv Health Sci Educ Theory Pract Date: 2012-03-28 Impact factor: 3.853