The purpose of this study was to verify that a laptop-based exercise in computational statistics allows the concept of multivariate analysis, the foundation of medical statistics, to be experienced manually, and that this exercise transforms the understanding of statistics and increases the value of computational statistics compared to learning to calculate up to a single regression by hand. Methods A survey of undergraduate Health Sciences students new to biomedical statistics was conducted using pre- and post-class questionnaires assessed in class over a 10year period. The statistical software used were R and RG (software that converts R commands to GUI). Results 98% of the students responded computer statistics　 when asked whether they felt that 　hand calculations　 or 　R and RG　 were more useful for future statistical analysis as an impression of multiple regression analysis. Discussion Simplifying data eliminates computational complexity, but at the same time it loses its realism. We believe that students' awareness of statistics will change if we strive to understand statistics from multiple perspectives through computer statistics exercises. 　We believe that a change in students awareness of statistics can be achieved not by simply making students understand statistics, but by making students understand statistics from multiple perspectives. Instead of making students understand statistics by a definition that prioritizes ease of manual computation, we believe that it can be achieved by making students understand statistics from multiple perspectives. A program that had students calculate computational statistics using simple data was considered to be an effective educational method for beginning students.

View full abstract