Multivariate Analysis

Multivariate analysis is a technique used mostly in social sciences to determine the relationships or effects of multiple variables simultaneously. Most often than not, the outcomes under scrutiny will be correlated and influenced by two or more predictors simultaneously.

There are two main multivariate distributions:

The type of multivariate analysis technique employed will depend on the nature of the sample data set used and the statistical objective. Some of the most common statistical techniques used by data scientist while carrying out multivariate analysis include:

Factor Analysis

This is essentially a data reduction technique. Data scientists mostly use this statistical technique in cases where the number of variables in the sample data set is large and they would wish to reduce them to a manageable number of factors.

This technique is further divided into two groups:

Multiple Regression Analysis

This is the most common multivariate technique. It is used in cases where you only have one response variable and multiple independent variables. The idea is to estimate the influence of each predictor on the criterion-variable when you hold the other predictors constant. The least squares method is utilized in this method.

Multiple Analysis of Variance (MANOVA)

This technique is used in cases where the data analyst has multiple categorical predictors and two or more outcome variables that are suspected to be correlated.