A statistical technique used to show how one dependent variable, such as sales, is affected by other variables, which are independent. Regression analysis measures how correlated the dependent and independent variables are. By using the correlation, an analyst can make an educated estimate about future events. For example, an office supply store may want to find out how the rate of sales depends on the unemployment rate or consumer spending.

Regression analysis is easily done in spreadsheet programs, and the results are presented in a scatter graph. A line connecting the two average variables is drawn to connect them to form the regression line. The regression line signifies a direct correlation between two variables, so if all the dots are connected, there is a direct correlation. The more the dots are dispersed outside the regression line, the less of a direct relationship there is between the two variables. The unexplained variation is called the coefficient of determination. Its square root is the correlation coefficient. If the coefficient is 1, there is a direct relationship between the variables. Regression analysis is used in risk-return analysis for portfolios and to analyze individual investments.