HOMEWORK -

The Linear Probability Model: Who smokes and who doesn't?

The EXCEL file firm-smoke_homework9 contains data from a survey on smoking behavior among employees in a large firm. Use the data provided, read the accompanying text file that explains the variables and their formats, upload in SPSS.

Then create a new DUMMY variable smoke defined as:

smoke = 0 iff cigs = 0

smoke = 1 iff cigs > 0

Make a summary statistics table including the following variables: smoke cigs age educ income restaurn price

Then run a simple linear probability regression with smoke as the dependent variable, to look for an answer to the following question:

Question 1) Is smoking related to the number of years of education? Explain the results in one sentence (maximum 20 words).

Question 2) After how many years of education does the model predict a negative probability that a worker smokes? Are you concerned about that?

Run a multiple linear probability regression with smoke as the dependent variable, and age and agesq100 as independent variables.

Question 3) At what age is a worker most likely to smoke?

Run a multiple linear probability regression with smoke as the dependent variable, and age agesq educ ln_income restaurn as independent variables.

Question 4) Interpret the ANOVA F-test. Conclude. Which explanatory variable(s) has/have no significant effect?

NOTE: the constant is NOT an explanatory variable (because it has no variation).

Drop the insignificant explanatory variable(s) from the regression, and add the variable ln_price as additional regressor to the model.

Question 5) What is the coefficient on ln_price, and how would you interpret it?

Question 6) Re-estimate the model with significant explanatory variables only. Then re-compute the answer for question 3. Explain the difference outcomes.

Attachment:- Assignment Files.rar