Correlation describes the strength, or degree, of linear relationship. That is, correlation lets us specify to what extent the two variables behave alike or vary together. Correlation analysis is used to assess the simultaneous variability of a collection of variables. The relationships among variables in a correlation analysis are generally not directional.

Example:

Suppose one wants to study the simultaneous changes with age of height and weight for a population. Then, one can assess the height and weight changes in the population from infants to adults. Regression analysis describes the how the change in height can influence the change in weight.

The measures of correlation infer the same thing as measures in the study of regression. Their value lies partly in their ability to measure the same thing in a slightly different way.

Correlation over time in which the level of response in one time period affects the level of response in the next period is called auto correlation. It is also called serial correlation.