Development and Evaluation of an Interactive Visual Workspace to Aid the Intuitive Understanding of ANOVA

Abstract

This report describes the development and evaluation of a Java-based interactive visual workspace, with accompanying tutorials, to help psychology undergraduates acquire an intuitive understanding of analysis of variance (ANOVA). The user is presented with two normal distributions and a sliding control allowing variation of the difference in group means, relative skewness and kurtosis. The distributions are generated algorithmically in real time to allow for a wide range of adjustments. An experimental group (n = 56) used the dynamic workspace while a control group (n = 52) studied a paper copy, both for 45 minutes. Performance on a 10 item multiple choice test was compared between groups, after no delay and after one hour, but, despite an overall mean correct score of 6.54 items, there were no significant condition effects. Subjective preference responses also did not discriminate between the groups. However, the cohort taking part in the evaluation study (i.e., the experimental and control groups combined) outperformed the cohort of the previous year on pertinent questions in the formal module examination. In conclusion, while there was no advantage to student learning of online interactive over static paper versions of the workspace, exposure to the materials created may have improved module examination performance. The research is continuing and the workspace is freely available from the Higher Education Academy Psychology Network website.