1. a set of statistical procedures for studying the relationships between one or more predictors and several outcome or dependent variables. Examples include canonical correlation analysis for assessing the relationships among two sets of variables; factor analysis for assessing the relationships among a large set of measures and a small set of underlying factors; multivariate analysis of variance and multivariate analysis of covariance for assessing potential group differences on several dependent variables; and structural equation modeling, which examines a theoretically based pattern of relationships among multiple independent, dependent, and even mediating variables. Also called multivariate statistics. Compare univariate analysis.

2. more generally, any procedure for understanding any large set of variables, whether dependent or independent.