1.5 pts Consider the following estimated Probit model: Pr(Yi 1lednuci, inci) (0.
ID: 3224144 • Letter: 1
Question
1.5 pts Consider the following estimated Probit model: Pr(Yi 1lednuci, inci) (0.05 0.03educi 0.67inci), where Y if individual i smokes and Yi 0 otherwise, educi indicates years of education, and inci denotes monthly income in thousands of dollars. The standard errors for the estimated coefficients are 0.021, 0.001, and 0.025, respectively (a) Interpret the sign of each coefficient. (b) Given the information at hand, how would you test the null that education does not affect the probability of smoking? Do you reject the null? (c) What is the marginal effect of inc on Pr(Y 1 educ inc) if educ 16 and inc 1.75? You just have to provide the formula, you do not need to provide the numerical valueExplanation / Answer
a)Probit regression, also called a probit model, is used to model dichotomous or binary outcome variables. In the probit model, the inverse standard normal distribution of the probability is modeled as a linear combination of the predictors.This can be interpreted, for example, that the a one unit change in the inc variable, increases the probability of smoking by 067. Similarly, for edu variable, decreases the probability of smoking by 0.03....
b) For testing purposes we can clearly see that education affect smoking habit in a decreasing way...so we will reject null hypothesis
c)marginal effect of income on smoking is 0.67*1.75
d) predicted probability is P(Y=1|edu,Inc)= capital phi of (0.05-0.03*16+0.67*1.75)
The numerical value can be obtaine from standard normal distribution table.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.