Purpose: To test whether there are effects of tutor expertise on student performance under conditions of curricular materials that have low or high levels of structure and that are poorly or well matched to students' levels of prior knowledge.
Method: The study was conducted in 1994-95 at the medical school of the University of Limburg. The data set used for analysis included 135 tutorial groups (with ten to 12 students per group), 119 tutors (each running only one group per unit), and 15 units in four curriculum years. The analysis was conducted at the level of tutorial groups since a tutor's level of expertise might differ for distinct units. Tutors were asked to judge their levels of expertise related to the cases discussed, based on which a distinction was made between expert and non-expert tutors. The degrees of structure of curricular materials and students' levels of prior knowledge were rated by the students. Using analyses of variance, students' scores on end-of-unit tests (each with about 150 true-false items) were compared for groups led by expert and non-expert tutors under conditions of low and high levels of structure and low and high levels of prior knowledge.
Results: No difference was found between the test scores of groups led by expert and non-expert tutors. The interaction effects between expertise and structure and expertise and prior knowledge also turned out to be not statistically significant.
Conclusion: The results suggest that expert tutors do not compensate for lack of curricular structure or students' lack of prior knowledge. This finding is not consistent with that of a recent study that expert tutors do compensate for lack of structure and lack of prior knowledge. This discrepancy may be accounted for by a much smaller range within which the structuredness of the curriculum and students' levels of prior knowledge varied in the present study compared with the previous study. An implication might be that faculty should put their efforts into designing structured curricula that are well matched to students' levels of prior knowledge instead of selecting hyper-expert tutors.