Multiple Regression and Marketing-Mix Models
This technical note provides a basic introduction to multiple linear regression. The concept of regression using a single independent variable is first introduced and then some of the practical challenges associated with it--including multiple independent variables in a regression--are discussed. Particular attention is paid to bias in the regression coefficients in the presence of omitted variables. The concept of the economic significance of a model is introduced and is contrasted with statistical significance. At Darden, it is used in a course elective titled "Big Data in Marketing."