Role of Control Variables in Regression
Control variables in regression are extra variables that are included in the model to remove the influence of other factors. This helps you understand the true relationship between the main independent variables and the dependent variable.
Most of the times other factors affect the dependent variable in regression model and leads to a misleading result. Control variable helps prevent misleading results by removing the influence of other factors. So, we will able to measure the real impact of our main predictors while holding other factors constant. Eventually, makes the regression results more reliable and closer to reality.
Example: Suppose we study Sales and advertising to predict or forecast of examine the effect of advertising on sales:
Sales = f(Advertising)
But sales also depend on Store Size, Customer Income If we don’t control these, we may wrongly think Advertising alone drives sales. So model should be
Sales = f(Advertising, Size, Income)
Here: Advertising is main predictor variable and Size, Income are the control variables.
Statlearner
Statlearner