14} What's the effect of being unemployed at some point in the past 10 years on the probability of being very happy? Use the probit model and determine this effect for someone with average values of educ, children, tvhours, and regattend. Round your answer to four decimal places. (6 points) 15} What's the marginal effect of another year of education on the probability of being very happy for someone who has 12 years of education, has never been unemployed, has no children, watches 1 hour of TV per day, and attends religious services regularly? Use the probit model to answer this question. Reund your nal answer to four decimal places. (6 points) 16) How do we evaluate model t in probit models? (Select all that apply) (4 points) 17) What's the effect of watching 5 more hours of TV per day on the probability of being very happy for someone who has never been unemployed, has 16 years of education, has no children, watches 1 hours of TV per day, and does not attend any religious services? Use the probit model to answer this question. Round the values of the betas to two decimal places. Round your nal answer to three decimal places. (6 points) 18) How would you test the hypothesis that the effect of previous unemployment on the proba bility of being very happy is offset by regular religious attendance? That is, if someone has been unemployed but attends religious services regularly, how do you test that the net effect on their probability of being very happy is zero? Write down the null and alternative hypotheses. Assume all values of the variables are held at their means. (5 points) 19) Suppose the pvalue=0.925 for the hypothesis test in question (18). What do you conclude? (5 points) 20} Which method does Stata use to estimate the probit model? (4 points) 21} Does watching more TV cause peOple to become less happy? Select the best answer. (5 points)