Select which statements are true about including a categorical confounding variable in a linear regression model: A. Including a categorical confounding variable in the model may change the measured effect of the explanatory variable on the response B. The coefficient is the difference in the slope of the explanatory variable associated with the confounding variable C. The categorical confounding variable operates like a slope term. For every one unit increases in the categorical confounder, the model predicts a corresponding change in the response variable that is the coefficient associated with the categorical variable D. The coefficient is the average difference in the outcome associated with levels of the confounder, after adjusting for the main explanatory variable