Consider the following regression results from a study about the impact of physician advice on consumption of alcohol (standard errors in parentheses; n = 500): DRINKS; = 13.0 + 11.4ADVICE; +2.8DIVSEP; +14.2UNEMP (2.12) (2.55) (5.16) # alcoholic drinks consumed by ith individual in last two weeks DRINKS ADVICE 1 if physician advised ith individual to reduce alcohol, 0 otherwise DIVSEP = 1 if ith individual divorced or separated, 0 otherwise UNEMPL 1 if ith individual unemployed, 0 otherwise = a) Which coefficient estimates are significant at conventional levels? b) Carefully state the meaning of the estimated coefficient for ADVICE. Does the sign make sense to you? What Classical Model assumption (other than Assumption ) is this equation almost surely violating? Explain. c) Could the type of problem you identified for ADVICE also apply in some way to DIVSEP or UNEMP? Explain.