Suppose we have a complicated model. Y and X are quantitative. Segment can be A, B, C, and Location can be domestic and international Model2 <- lm(log(Y) ~ X + Segment + Location) We obtain the coefficients table below: Variable Estimate of coefficient p-value (intercept) -3 X 2 Less than 0.0001 Segment B 0.4 Less than 0.0001 Segment C -1.3 Less than 0.0001 International 3 Less than 0.0001 Interpretations of some of the dummy variables: Interpretation computed from the 0.4 coefficient of Segment B: Since e^(0.4) = 1.49, we can say "When we control for X and Location, the predicted Y in Segment B is 1.49 times higher in Segment B than Segment A." Since (1.49-1)*100% = 49%, we can say "When we control for X and location, the predicted Y in Segment B is 49% higher in Segment B than Segment A." Interpretation computed from the -1.3 coefficient of Segment C. Since the coefficient is negative, it's best to use a percentage change e-1.3 = 0.27 and (0.27-1)*100% = -73% So "When we control for X and location, the predicted Y in Segment C is 73% less in Segment C than Segment A. Module 6.5 Question 1 HomeworkUnansweredDue Today