Multivariate Analysis of Covariance (MANCOVA)
Multivariate Analysis of Covariance · Also known as: MANCOVA, multivariate ANCOVA, MANOVA with covariates, MANCOVA — Çok Değişkenli Kovaryans Analizi
MANCOVA (Multivariate Analysis of Covariance) is a parametric hypothesis test that simultaneously compares two or more groups on multiple continuous dependent variables while statistically controlling for one or more covariates. It extends MANOVA by incorporating covariate adjustment, a tradition consolidated in multivariate statistical methodology by the 1970s and authoritatively documented by Tabachnick and Fidell (2019).
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use MANCOVA when you have two or more independent groups, two or more continuous dependent variables measured on all participants, and one or more continuous covariates you want to statistically control. The method requires: multivariate normality of the dependent variables within each group, homogeneity of variance-covariance matrices across groups (tested with Box's M), a linear relationship between each covariate and each dependent variable, and homogeneity of regression slopes — meaning the covariate-outcome relationship does not differ across groups (the parallel slopes assumption). A minimum sample of approximately 50 observations is recommended, with more required as the number of dependent variables and covariates grows. If multivariate normality is severely violated, consider bootstrapped MANCOVA or nonparametric alternatives.
Strengths & limitations
- Controls experiment-wise Type I error by testing all dependent variables simultaneously rather than running separate ANCOVAs.
- Removes covariate-driven noise from group comparisons, increasing statistical power and precision.
- Detects group differences that may not be apparent in any single dependent variable but exist as a multivariate pattern.
- Produces a single omnibus test that is conceptually cleaner than multiple univariate tests when outcomes are correlated.
- Requires the homogeneity of regression slopes assumption; if slopes differ across groups, the MANCOVA adjustment is misleading.
- Sensitive to violations of multivariate normality and variance-covariance homogeneity, particularly with small or unequal group sizes.
- Results can be difficult to interpret when many dependent variables are included and the multivariate effect is significant but univariate follow-ups are mixed.
- Requires larger samples than MANOVA because the additional covariate parameters consume degrees of freedom.
Frequently asked
What is the difference between MANCOVA and MANOVA?
MANOVA tests whether groups differ on a combination of dependent variables without any covariate control. MANCOVA does the same but first partials out the linear influence of one or more continuous covariates, making the group comparison cleaner when participants differ on background characteristics that predict the outcomes.
Which multivariate statistic should I report — Wilks' Lambda, Pillai's Trace, or another?
Wilks' Lambda is the most commonly reported statistic and is appropriate when assumptions are met. Pillai's Trace is the most robust to violations of multivariate normality and variance-covariance homogeneity, making it preferred when sample sizes are small or unequal. Roy's Largest Root is the most powerful but only when the group differences are concentrated on a single variate. Tabachnick and Fidell (2019) recommend Pillai's Trace as the default choice in applied work.
How do I check the homogeneity of regression slopes assumption?
Add the interaction between each covariate and the grouping variable to the model. If any covariate-by-group interaction is statistically significant, the slopes are non-homogeneous and the standard MANCOVA adjustment is invalid. In that case, consider moderated regression or separate within-group analyses.
What follow-up analyses should I run after a significant MANCOVA?
Conduct separate univariate ANCOVAs for each dependent variable using a Bonferroni-corrected alpha (e.g., 0.05 / number of DVs) to identify which outcomes drive the overall multivariate effect. Report partial eta-squared for each effect. Do not run multiple post-hoc pairwise comparisons without alpha correction.
Sources
- Tabachnick, B. G. & Fidell, L. S. (2019). Using Multivariate Statistics (7th ed.). Pearson. ISBN: 978-0134790541
How to cite this page
ScholarGate. (2026, June 1). Multivariate Analysis of Covariance. ScholarGate. https://scholargate.app/en/statistics/mancova
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- ANCOVAStatistics↔ compare
- Discriminant AnalysisStatistics↔ compare
- Hotelling's T² TestStatistics↔ compare
- MANOVAStatistics↔ compare
- One-way ANOVAStatistics↔ compare