You decide to dig deeper and obtain the following numbers (Y is the revenue generated in CHF, D is an indicator which is one of the employee participated in the seminar, and zero otherwise). Variable Average in subsample Average in subsample D=0 D=1 65'636 CHF 70'693 CHF Y D 0 1 X2 sales revenues (CHF) =1 if participated in seminar, =0 else age of sales representative tenure of representative with the 42 44 X3 16 21 X4 0.73 0.65 X5 company =1 if male, =0 if female revenue generated in the last year (CHF) 49'497 CHF 59'764 CHF (a) Interpret the numbers in the table in the context of your identifying assumptions made so far. Do you think that the two subsamples are balanced? What information do you miss to empirically assess this question? (3 points) (b) Your colleague suggests to estimate the ATE of seminar participation using the specification Y = a + B1D + and to calculate the estimator , with OLS. Formally show that the OLS estimator gives the same result as the difference-in-means estimator. (10 points)