MANOVA ("multiple analysis of variance") is a procedure for testing the equality of mean vectors of more than two populations. The technique is analogous to ANOVA for univariate data, except that groups are compared on multiple response variables simultaneously. While -tests can be used in the uniseriate case to assess the hypothesis under consideration, there is no single test statistic in the multivariate case that is optimal in all situations (Everitt and Wykes 1999, p. 125).

REFERENCES:

Bijleveld, C. C. J. H.; van der Kamp, L. J. T.; Mooijaart, A.; van der Kloot, W. A.; van der Leeden, R.; and van der Burg, E. Longitudinal Data Analysis: Designs, Models and Methods. London: Sage, 1998.

Everitt, B. S. and Wykes, T. Dictionary of Statistics for Psychologists. London: Arnold, p. 125, 1999.