Analysis of Correlation and Regression

It is quite often that one is interested to quantify the dependence (positive or negative) between two or more random variables. The basic role of covarianceCovariance is to identify the nature of dependence. However, the covariance is not an appropriate measure of dependence since it is dependent on the scale of observations. Hence, a measure is required which is unaffected by such scale changes. This leads to a new measure known as the correlation coefficient. Correlation analysis is the study of analyzing the strength of such dependence between the two random variables using the correlation coefficient. For instance, if X represents the age of a used mobile phone and Y represents the retail book value of the mobile phone, we would expect smaller values of X to correspond to larger values of Y and vice versa.