Question 6

smoking.csv file: https://docs.google.com/spreadsheets/d/1yR3qHfaTBq9e6lqbRqfmupkzJFbiffbXVOpraEr-934/edit?usp=sharing

We use the dataset "smoking.csv" that we analysed in Tutorial 09 We are interested in estimating the smoking probability as a function of age For all estimation of linear probability models, the type of standard error should be "HC1" or equivalently "slam" For the questions below, it is assumed that the required Repackages are loaded and the data are attached, QUESTION 6 Regress smoker on quadratic polynomials of age, using a probit regression. What is the maximum of the fitted (predicted) probabilities in the data? Report your answer (a value between 0 and 1) to two decimal places. Hint: In a regression of Yi on X; by a probit regression, the fitted probability at X; is given as Pr ( Y, = 1/ x, ) = $ ( Bo + BIX) where b(z) is the CDF of N(0, 1). > help(gls) might also be useful