MVN: An R Package for Assessing Multivariate Normality

Abstract

Assessing the assumption of multivariate normality is required by many parametric multivariate statistical methods, such as MANOVA, linear discriminant analysis, principal component analysis, canonical correlation, etc. It is important to assess multivariate normality in order to proceed with such statistical methods. There are many analytical methods proposed for checking multivariate normality. However, deciding which method to use is a challenging process, since each method may give different results under certain conditions. Hence, we may say that there is no best method, which is valid under any condition, for normality checking. In addition to numerical results, it is very useful to use graphical methods to decide on multivariate normality. Combining the numerical results from several methods with graphical approaches can be useful and provide more reliable decisions. Here, we present an R package, MVN, to assess multivariate normality. It contains the three most widely used multivariate normality tests, including Mardia's, Henze-Zirkler's and Royston's, and graphical approaches, including chi-square Q-Q, perspective and contour plots. It also includes two multivariate outlier detection methods, which are based on robust Mahalanobis distances. Moreover, this package offers functions to check the univariate normality of marginal distributions through both tests and plots. Furthermore, especially for non-R users, we provide a user-friendly web application of the package. This application is available at

Applicability of Demirjian’s four methods and Willems method for age estimation in a sample of Turkish children

“…residuals, which is the magnitude of difference between estimated age and chronological age. The response variable was symmetrically distributed among comparison groups, however; statistical tests rejected normality as a result of large sample size [36,37]. Previous studies showed that ANOVA is robust to moderate departures from normality and can be used when sample size is large enough [38,39].…”

Phenotypic plasticity in sex pheromone production in Bicyclus anynana butterflies

“…Multivariate outlier detection was done based on robust Mahalanobis distance. All analyses were carried out using the R language and environment for statistical graphics and computing63, and associated packages ( MVN v.4.064 and car 65). Strength of shift in MSP amounts was calculated as standardized mean differences.…”

Apical dominance ratio as an indicator of the growth conditions favouring Abies alba natural regeneration under Mediterranean environment

“…In particular, the multivariate outliers were graphically checked by the QQPlot (''chisplott'' function) for the observed Mahalanobis distances. Bartlett's test was used to test the homogeneity of variance (homoscedasticity), while either univariate and multivariate normality were performed with the Shapiro-Wilk and Mardia skewness tests (''MVN'' R package by Korkmaz et al 2014), respectively. In order to investigate which variables account for differences among ADR classes, univariate ANOVA tests were also performed.…”