MANOVA Post Hoc Techniques Used in Published Articles: A Systematic Review

Abstract

Researchers are usually interested in examining group mean differences after exhibiting significant MANOVA test. The MANOVA-ANOVA approach has been recognized as an issue in several previous research reviews. The enormous growth of social and behavioral research in the past decade has necessitated the need for and use of research reviews. Therefore, the purpose of this review is to identify trends in data-analytic practices in pertain to MANOVA follow-up procedures and to examine what specific tests have been employed. Articles from five major databases (e.g. ERIC, PsychINFO) that regularly publish research in education were examined. The results of analyzing 235 studies were reported. The results indicated that to explain significant MANOVA results, univariate tests were the most post hoc method applied in 95% of the studies, whereas multivariate methods such as DFA was only applied in 5% of the studies.