MULTIVARIATE ANALYSIS ASSIGNMENT HELP

What is the difference between bivariate and multivariate analysis techniques?

In bivariate analysis, a data scientist considers the effects of independent variables on only one dependent variable. On the other hand, multivariate analysis is concerned with determining the effects of the independent variables on more than one dependent variable.

Most problems in social statistics are described using multivariate analysis. This is due to the hypothesis by social scientists that various real-life outcomes or decisions can be described by multiple responses.

Multivariate analysis encompasses multiple data analysis techniques. All these techniques are based on the use of a combination of descriptive statistics such as means, correlations, covariances and variances. The choice of multivariate analysis technique to employ will majorly depend on two things;

1. The main research questions the analyst seeks to answer
2. The assumptions made by the analyst about the sample data. This include concerns such as whether the data has equal variance and whether it is linear and normally distributed. An examination of measures such as kurtosis and skewness gives an understanding of the distribution.

NOTE: Multivariate analysis is most effective when you have a very large data set. A small sample data with outliers will lead to misleading conclusions. Missing values on a small data sample will also yield the same result.

Examples of multivariate analysis techniques include:

1. Logistic regression
2. Multidimensional scaling
3. Factor analysis
4. Discriminant analysis
5. Multivariate analysis of variance (MANOVA)
6. Principal-components analysis
7. Path analysis
8. Multiple regression analysis