Linear regression is a statistical programming method where the score of a variable ‘A’ is predicted from the score of a second variable ‘B’. B is referred to as the predictor variable and A as the criterion variable.