Learning Outcomes in a Laboratory Environment vs. Classroom for Statistics Instruction: An Alternative Approach Using Statistical Software

Abstract

The role of any statistics course is to increase the understanding and comprehension of statistical concepts and those goals can be achieved via both theoretical instruction and statistical software training.  However, many introductory courses either forego advanced software usage, or leave its use to the student as a peripheral activity.  The purpose of this study was to determine if there was instructional value in replacing classroom time with laboratory time dedicated to statistical software usage. The first approach used classroom lecture presentations, while the second replaced one classroom period per week with statistical software laboratories.  It was hypothesized that replacing classroom time with software based laboratories would increase the level of statistics knowledge as compared to an otherwise identical class with no lab based component.   Both pre-course and end-of course surveys were used, as well as identical examination questions.  Comparisons within a time point, and longitudinal performance over the course were both evaluated.  Survey results indicated that students would recommend lab based instruction significantly more than a primarily lecture based instruction (32% more, p=.020).  Additionally, the performance improvement over the course of the semester was significantly higher for those students participating in laboratories (19.2% increase, p=.011).  These findings indicate that sacrificing classroom time for a laboratory period improves the educational experience in an introductory statistics course and may help with the understanding and retention of difficult topics.