Effects of a learning-centered approach to assessment on students' attitudes towards and knowledge of statistics

Objective To examine the effects of learning-centered assessment on students' attitudes toward and knowledge of statistics. Methods This nonequivalent control group study examined second-year (P2) and third-year (P3) year PharmD students enrolled in a required course on statistics. P2 students received a learning-centered assessment approach; P3 students received an instruction-centered approach. Course topics, sequencing, and examinations were identical between the courses. Students' statistical self-efficacy, attitudes toward statistics, statistical knowledge, and course perceptions were measured at the beginning and end of the course. Results Statistical self-efficacy increased over time but was not statistically different between classes. Statistical knowledge increased significantly ( p < 0.001) over the semester, with greater increases for P2 students than P3s. A significant improvement in statistics attitudes was noted in the SATS-36 difficulty domain ( p = 0.026) over the semester, but this was not different between classes. Compared with P3s (instruction-centered), the P2 students (learning-centered) had significantly better perceptions regarding the course with respect to having multiple opportunities to demonstrate learning ( p < 0.001), control in the determination of the course grade ( p < 0.001), focusing on learning rather than course grade ( p = 0.033), and receiving adequate feedback ( p = 0.006). Conclusions There were changes in statistical self-efficacy and attitudes toward statistics among the students, but these appeared to be the effect of time rather than the assessment approach. Students in the learning-centered course had greater increases in statistical knowledge and more positive perceptions of the learning environment. More research comparing learning-centered teaching with traditional approaches is needed.