You are on page 1of 3

Assignment STATA 1

AHMED SALEEM KHAN


18U03086
SEC A
a. Write down the command that you will use to verify the total number of observations
stated above.
Command: count
9822
b. What fraction of the sample is employed at the time of the interview?
Command: Tab employ
89.22% are employed

c. What fraction of the sample has abused alcohol?

Command: Tab abuse


9.92% have abused alcohol

d. Run a simple regression of employ on abuse. Is the relationship as you expected? Is it


statistically significant?

Command: reg employ abuse


The regression results show a negative relationship between employ and abuse (as
expected), indicating that an individual who abuses alcohol has a 2.83% lower chance of
being employed as compared to someone who does not, ceteris paribus. This result is
statistically significant at the 5% level as p = 0.006

e. Obtain the fitted values for the LPM estimated in part (d) and report what they are when abuse
=0and when abuse=1.

Command: predict y hat

When abuse = 0, fitted value = 0.9009946

When abuse = 1, fitted value = 0.87269

f. Run a probit of employ on abuse. Do you get the same sign and statistical significance as in
part(d)? Command: probit employ abuse

We get the same sign as the LMP model as well as the same significance but a different slope
value. The abuse coefficient is still significant at the 5% level.
g. Obtain the fitted values for the probit model estimated in part (f).

Command: predict probit y hat

When abuse = 0, fitted value = 0.9009946

When abuse = 1, fitted value = 0.87269

h. Run a logit of employ on abuse.

Command: logit employ abuse

i. Obtain the fitted values for the logit model estimated in part (h).
Command: predict logit y hat
When abuse = 0, fitted value = 0.9009946
When abuse = 1, fitted value = 0.87269

Question 2: Use the data in LOANAPP.RAW for this exercise. a. Estimate a probit model of
approve on white. Find the estimated probability of loan approval for both whites and nonwhites.
How do these compare with the linear probability model estimates?

Command: probit approve white

Predict y hat probit

The probability of a loan approval for a white person is 0.908 while the probability of a loan
approval for a non-white person is 0.707, ceteris paribus.

Command: reg approve white

Predict y hat

The LPM model gives the same fitted values as the probit model

c. Now, add the variables hrat, obrat, loanprc, unem, male, married, dep, sch, cosign, chist,
pubrec, mortlat1, mortlat2, and vr to the probit model. Is there statistically significant
evidence of discrimination against nonwhites?
Command: probit approve white hrat obrat loanprc unem male married dep sch cosign
chist pubrec mortlat1 mortlat2 vr
According to this regression, the variable white is statistically significant (0<0.005)
suggesting there is some level of discrimination against non-whites.
d. Estimate the model from part (b) by logit. Compare the coefficient on white to the probit
estimate.
Command: logit approve white hrat obrat loanprc unem male married dep sch cosign
chist pubrec mortlat1 mortlat2 vr
The coefficient on white for the logit model is much higher than that of the probit model.
The probit model coefficient was 0.5202525 while the logit model has a coefficient of
0.9377643 and the logit coefficient is also significant.

You might also like