A comparison of two follow-up analyses after multiple analysis of variance, analysis of variance, and descriptive discriminant analysis: A case study of the program effects on education-abroad programs

Abstract

This study compared the utility of two different post-hoc tests after detecting significant differences within factors on multiple dependent variables using multivariate analysis of variance (MANOVA). We compared the univariate F test (the Scheffé method) to descriptive discriminant analysis (DDA) using an educational-tour survey of university study-abroad students. Differences on five dependent measures of study-abroad program outcomes were compared by students' academic major and in-country residency arrangements (residence hall, home stay, apartment). Univariate F test and DDA were used to detect group differences in the dependent variables after the overall significant F test in MANOVA. The results indicated that the two post-hoc tests were congruent. However, DDA served better than the univariate F test to determine which outcome variables contributed most to separating the independent variables. We conclude that both post-hoc analyses should be utilized after a significant MANOVA to obtain accurate and full insight into the data.

Parent Publication

• Proceedings of the 2009 Northeastern Recreation Research Symposium

Citation

Yu, Alvin H.; Chick, Garry. 2010. A comparison of two follow-up analyses after multiple analysis of variance, analysis of variance, and descriptive discriminant analysis: A case study of the program effects on education-abroad programs. In: Watts, Clifton E., Jr.; Fisher, Cherie LeBlanc, eds. Proceedings of the 2009 Northeastern Recreation Research Symposium. Gen. Tech. Rep. NRS-P-66. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station: 235-240.