Sei sulla pagina 1di 5

Econometrics 206-1

Exam III: 10.10 AM -11.40 AM, 24 April 2017

In answering these below, paste the Stata output only when it is asked. When
pasting output, use the copy as picture option. When testing a hypothesis, be sure
to mention the distribution of the test statistic, its degrees of freedom, the level of
significance and the associated critical value. DO NOT USE THE STATA test
COMMAND.

It would be easiest if you inserted your answer between the questions below and
returned this document. Rename the document as `your name.docx’ and upload it
on LMS.

You have to do this exam by yourself. You are allowed to consult the textbook and
your notes. You are NOT allowed to consult anybody whether by speaking, by text
messages or email or any other means. Violations will attract penalties as per
Ashoka policy.

1. (a) Regress log of wages on a constant and the female dummy. Paste output
here.
. regress logwages female

Source SS df MS Number of obs = 1,000


F(1, 998) = 181.25
Model 140.049853 1 140.049853 Prob > F = 0.0000
Residual 771.131872 998 .772677226 R-squared = 0.1537
Adj R-squared = 0.1529
Total 911.181724 999 .912093818 Root MSE = .87902

logwages Coef. Std. Err. t P>|t| [95% Conf. Interval]

female -.8553218 .0635312 -13.46 0.000 -.9799919 -.7306516


_cons 4.12635 .0322699 127.87 0.000 4.063026 4.189675

(b) Interpret the coefficient on the female dummy.


Coefficient of the female dummy in the regression is -.855. This implies that if the
participant is a female then the value of wages decreases by 85.5% points.

(c) Test the null hypothesis that the coefficient on female dummy is -0.5 against
the alternative that the coefficient on female dummy is less than -0.5. Show your
workings.
Here,
Let the coefficient of the female dummy be BetaHat1.
Null hypothesis: Hnot: BetaHat1=-0.5
Alternate hypothesis: Ha: BetaHat1 < -0.5

T = (BetaHat1 – (-0.5))/s.e.(BetaHat1) = -5.59 where s.e. is the standard error


Now, at both significance level, it is a one-tail test,
Since t < -1.96 (at 5% significance), we reject the null hypothesis and accept the
alternate hypothesis that Beta1 is less than -0.5 at 5 % significance level.

Similarly, since t <- 3.84 (at 1 % significance), we reject the null hypothesis and
accept the alternate hypothesis that Beta1 is less than -0.5 at 1% significance level.

[5+5+10]

2. (a) Regress log of wages on a constant, the female dummy, age of the
individual and the square of age. Paste your output here.
. regress logwages female age agesq

Source SS df MS Number of obs = 1,000


F(3, 996) = 95.62
Model 203.744072 3 67.9146908 Prob > F = 0.0000
Residual 707.437652 996 .710278767 R-squared = 0.2236
Adj R-squared = 0.2213
Total 911.181724 999 .912093818 Root MSE = .84278

logwages Coef. Std. Err. t P>|t| [95% Conf. Interval]

female -.8576719 .0609852 -14.06 0.000 -.9773461 -.7379977


age .0634126 .0103899 6.10 0.000 .0430241 .0838011
agesq -.0006079 .0001343 -4.53 0.000 -.0008714 -.0003445
_cons 2.748186 .1880344 14.62 0.000 2.379197 3.117175

(b) Controlling for age and the square of age does not seem to substantially
change the coefficient of the female dummy. Why is that so?

The coefficient of the female dummy after controlling for age and the square of
age is -0.858 whereas before controlling for age and the square of age was -
0.855. Controlling for age and the square of age does not seem to substantially
change the coefficient of the female dummy. This is because there is hardly any
collinearity between the age (for this matter even the square of age) and the
female dummy variable. Had the independent variables be correlated, the
coefficient of the female dummy would have changed significantly.
[5+5]

3. (a) Regress log of wages on a constant, the female dummy, age of the
individual the square of age and the social group dummies for scheduled caste,
for scheduled tribe and for other backward caste. Note the omitted category is
the general castes (or forward castes). Paste your output here.
. regress logwages female age agesq scd std obc

Source SS df MS Number of obs = 1,000


F(6, 993) = 56.37
Model 231.515509 6 38.5859182 Prob > F = 0.0000
Residual 679.666215 993 .684457417 R-squared = 0.2541
Adj R-squared = 0.2496
Total 911.181724 999 .912093818 Root MSE = .82732

logwages Coef. Std. Err. t P>|t| [95% Conf. Interval]

female -.807476 .0605312 -13.34 0.000 -.9262597 -.6886923


age .0605878 .0102103 5.93 0.000 .0405516 .0806241
agesq -.0005791 .0001319 -4.39 0.000 -.0008379 -.0003202
scd -.4413509 .0755029 -5.85 0.000 -.5895144 -.2931873
std -.2481139 .0887235 -2.80 0.005 -.4222209 -.0740068
obc -.3809426 .0708538 -5.38 0.000 -.519983 -.2419022
_cons 3.085829 .1928521 16.00 0.000 2.707385 3.464274

(b) Test the null hypothesis that none of the social group dummmies matter, i.e.,
controlling for sex, age and square of age, the average of log wages is the same
for all categories: scheduled castes, scheduled tribes, other backward castes and
the general (forward) castes. Do NOT use the Stata test command.

Null Hypothesis – H0: Beta4 = 0 , Beta5 = 0, Beta6 = 0


Alternate hypothesis – H1: H0 is not true.
To test this hypothesis, we need to conduct a F test. Where the F static is given by-
F = ((SSR(r) – SSR(ur))q)/SSR(ur) / (n-k-1)
F = 13.534

The number is well above the 1% critical level for a f test with 993 degrees of
freedom. So, we reject the null hypothesis.
(c) Test the null hypothesis that relative to the general (forward) castes, scheduled
castes and other backward castes suffer the same extent of discrimination. If this
requires new regressions, paste the output in your answer.

Null Hypothesis - H0 : Beta4 = Beta6 or Beta4 – Beta6 = 0


Alternate Hypothesis – H1: Beta4 – Beta6 not= 0
The P value for this test is 0.3673. That means that we fail to reject H0 even at very
high levels of Significance.
[5+15+15]

4. (a) Regress log of wages on a constant, the female dummy, age of the
individual, the square of age, the social group dummies for scheduled caste, for
scheduled tribe and for other backward caste, and the education dummies for
illiterate, literate, primary, secondary, and higher secondary. Paste the output
here.
. regress logwages female age agesq scd std obc illiterate literate primary secon
> dary higher_secondary

Source SS df MS Number of obs = 1,000


F(11, 988) = 85.16
Model 443.473774 11 40.3157976 Prob > F = 0.0000
Residual 467.707951 988 .473388614 R-squared = 0.4867
Adj R-squared = 0.4810
Total 911.181724 999 .912093818 Root MSE = .68803

logwages Coef. Std. Err. t P>|t| [95% Conf. Interval]

female -.6287476 .0524421 -11.99 0.000 -.7316584 -.5258369


age .0434705 .0086153 5.05 0.000 .026564 .0603769
agesq -.0003424 .0001108 -3.09 0.002 -.0005599 -.000125
scd -.2332977 .0636376 -3.67 0.000 -.3581781 -.1084173
std -.0939275 .0744076 -1.26 0.207 -.2399427 .0520876
obc -.2183765 .0596753 -3.66 0.000 -.3354813 -.1012716
illiterate -1.579581 .082054 -19.25 0.000 -1.740601 -1.418561
literate -1.302972 .0962819 -13.53 0.000 -1.491913 -1.114032
primary -1.303007 .0929865 -14.01 0.000 -1.485481 -1.120534
secondary -.9743013 .0859749 -11.33 0.000 -1.143016 -.8055869
higher_secon~y -.4865914 .1186415 -4.10 0.000 -.7194097 -.253773
_cons 4.334187 .1815994 23.87 0.000 3.977822 4.690552

(b) Compare the above regression with the regression in question 3 (without the
education dummies). Does the inclusion of education dummies alter the
discrimination against women, scheduled castes, scheduled tribes and other
backward castes? Why?
The inclusion of the education dummies in the regression does alter the
estimates of discrimination against women, scheduled castes, scheduled tribes
and other backward castes. This is because the inclusion of education dummies
has a positive impact to reduce the extent of discrimination against women,
scheduled castes, scheduled tribes and other backward castes. Further, there is
some collinearity between the education dummies and the discrimination
against women, scheduled castes, scheduled tribes and other backward castes.
Because of the collinearity between the education dummies and the social group
dummies, the inclusion of education dummies alter the discrimination against
women, scheduled castes, scheduled tribes and other backwarded castes.
[5+15]

5. (a) To the explanatory variables in the regression in Qn 4(a), add land owned
(LandO) and land possessed (LandP) and re-run the regression. DO NOT paste
the output.

For Land0, standard error = 0.0000767


For LandP, standard error = 0.0000769

(b) Is either of the land variables individually significant at the 5 or 10% level?
Neither of the land variables are individually significant at the 5% level or 10%
level.

(c) Now drop land owned (LandO) and re-run the regression. Is the included
land variable significant at the 5 or 10% level?

After dropping LandO, standard error for LandP = 0.0000177

After dropping the LandO variable from the regression, LandP variable is
significant at both 5% and 10% level.

(c) Explain the pattern of results observed in (b) and (c).


There is a linear relationship between land owned and land possessed. Land
possessed includes land owned as well. So, when including both land owned and
land possessed variable, they fail to make a significant impact as explanatory
variables at 5% and 10% significance level. Whereas, after dropping land owned
variable, land possessed is significant at both 5% and 10% level. This pattern is
because of the linear relationship between the land owned and land possessed.
[4+4+7]

Potrebbero piacerti anche