You are on page 1of 10

EES 404: ECONOMIC DATA ANALYSIS $ COMPUTER SIMULATION

MARKING GUIDE

SEM II 2022/2023

Question One

a) Explain the difference between time series and cross- sectional data (4 Marks)

b) Your presented with the following wage model


2 2
Wage=30,000+0.45 exp+ 0.58 Educ +2.05 Age−0.12 Age Adj . R =0.53
(10,000) (0.1) (0.32) (2.1) (0.05)
Required
i) Test the significance of the coefficients and interpret the results (12 Marks)

30000 Reject the null/ Significant


Constant term =3
10000
0.45 Reject the null/ Significant
Experience =4.5
0.1
0.58 Don’t reject the null/ Not Significant
Education =1.8
0.32
2.05 Don’t reject the null/ Not Significant
Age =0.98
2.1
0.12 Reject the null/ Significant
Age2 =2.4
0.05

ii) Explain the relationship between Wage and Age (5 Marks)


Nonlinear
iii) Explain a simple test that you would use to check for heteroskedasticity (5 Marks)
 Run the regression
 Estimate the error term
 Plot the error tern against the dependent variable

Prepared by Dr. Makambi


iv) How much will an individual aged 45 years, educated for 16 years with 5 years of
experience earn? (4 Marks)
2
30,000+0.45 × 5+0.58 ×16+2.05 × 45+0.12 × 45 =KES 29,860

QUESTION TWO

a) Explain the meaning of endogeneity and cite the main channels through which the problem
arises (6 Marks)
b) Evaluate Table I, II, III & IV and answer the following questions
Table I
Instrumental variables (2SLS) regression Number of obs = 722
Wald chi2(5) = 102.33
Prob > chi2 = 0.0000
R-squared = 0.1412
Root MSE = .38838

lwage Coef. Std. Err. z P>|z| [95% Conf. Interval]

educ .1166336 .0181083 6.44 0.000 .081142 .1521252


exper .0277163 .0056976 4.86 0.000 .0165492 .0388835
1.married .1868934 .0462632 4.04 0.000 .0962193 .2775675
tenure .0071768 .0031142 2.30 0.021 .0010731 .0132805
1.black -.1166315 .0539041 -2.16 0.030 -.2222816 -.0109814
_cons 4.68515 .3002695 15.60 0.000 4.096633 5.273668

Instrumented: educ
Instruments: exper 1.married tenure 1.black sibs meduc feduc hours hourssq

i) Identify the endogenous variable and instruments used (2 Marks)

Table II

. estat endog

Tests of endogeneity
Ho: variables are exogenous

Durbin (score) chi2(1) = 7.79709 (p = 0.0052)


Wu-Hausman F(1,715) = 7.80579 (p = 0.0053)

Table III

Prepared by Dr. Makambi


. estat firststage

First-stage regression summary statistics

Adjusted Partial
Variable R-sq. R-sq. R-sq. F(5,712) Prob > F

educ 0.3471 0.3388 0.1634 27.8109 0.0000

Table IV

. estat overid

Tests of overidentifying restrictions:

Sargan (score) chi2(2) = .97248 (p = 0.6149)


Basmann chi2(2) = .960304 (p = 0.6187)

ii) Evaluate the diagnostics presented in tables II through IV and comment on whether Two Stage
Least Squares Estimator is appropriate (6 Marks)

Question Three

Results showing the determinants of fertility rate (measured using number of living children)

Table I

Prepared by Dr. Makambi


. reg children age educ heduc agefm i.bicycle i.electric i.usemeth urban, vce(robust)

Linear regression Number of obs = 1,905


F(8, 1896) = 191.26
Prob > F = 0.0000
R-squared = 0.4703
Root MSE = 1.6579

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

age .1807723 .0064139 28.18 0.000 .1681933 .1933512


educ -.0896077 .0118793 -7.54 0.000 -.1129055 -.0663099
heduc -.040427 .010247 -3.95 0.000 -.0605235 -.0203304
agefm -.0702758 .0098997 -7.10 0.000 -.0896912 -.0508604
1.bicycle .3929862 .0823187 4.77 0.000 .2315414 .554431
1.electric -.3936685 .1126045 -3.50 0.000 -.6145103 -.1728267
1.usemeth 1.088764 .0905267 12.03 0.000 .9112214 1.266306
urban -.2752669 .0823493 -3.34 0.001 -.4367717 -.1137621
_cons -.8280882 .2116453 -3.91 0.000 -1.24317 -.413006
Residuals

-10
10

-5
5

Figure I

0 1000 2000 3000 4000


ID

Prepared by Dr. Makambi


Figure II
15

10

10 20 30 40 50 0 5 10 15 20
age in years years of education

number of living children Fitted values number of living children Fitted values

Figure III

Prepared by Dr. Makambi


Residuals

Residuals
-10

-10
10

10
-5

-5
5

0
10 20 30 40 50 10 20 30 40 50
age in years age at first marriage

0 5 10 15 20 0 5 10 15 20
years of education husband's years of education

Figure IV

Prepared by Dr. Makambi


Density
.3

.2

.1

0
Kernel density estimate

-10 -5 0 5 10
Residuals

Kernel density estimate


Normal density
kernel = epanechnikov, bandwidth = 0.2818

Figure V

Prepared by Dr. Makambi


Normal F[(r-m)
1.00

0.75

0.50

0.25

0.00
0.00 0.25 0.50 0.75 1.00 -5 0 5
Empirical P[i] = i/(N+1) Inverse Normal

Normal F[(r-m)/s] Reference Residuals Reference

Table II

. hettest

Breusch-Pagan / Cook-Weisberg test for heteroskedasticity


Ho: Constant variance
Variables: fitted values of children

chi2(1) = 296.22
Prob > chi2 = 0.0000

Table III

. ovtest

Ramsey RESET test using powers of the fitted values of children


Ho: model has no omitted variables
F(3, 1893) = 2.41
Prob > F = 0.0650

Prepared by Dr. Makambi


Table IV

. linktest

Source SS df MS Number of obs = 1,905


F(2, 1902) = 844.64
Model 4627.56918 2 2313.78459 Prob > F = 0.0000
Residual 5210.31848 1,902 2.73938932 R-squared = 0.4704
Adj R-squared = 0.4698
Total 9837.88766 1,904 5.16695781 Root MSE = 1.6551

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

_hat 1.059925 .096777 10.95 0.000 .870125 1.249726


_hatsq -.0084299 .0131766 -0.64 0.522 -.034272 .0174122
_cons -.0858861 .1625149 -0.53 0.597 -.4046123 .2328401

Table V

. vif

Variable VIF 1/VIF

age 1.26 0.796617


educ 2.02 0.494700
heduc 1.93 0.518183
1.usemeth 1.13 0.887546
1.electric 1.75 0.570532
1.tv 1.77 0.565864
1.bicycle 1.02 0.980953
1.urban 1.23 0.814046
agefm 1.16 0.858413

Mean VIF 1.47

Required

Identify the appropriate figure(s) and/or table(s) and use the outputs to evaluate the following
regression diagnostics and regression model estimates

i) Normality (3 Marks)
ii) Heteroskedasticity (4 Marks)
iii) Linearity (3 Marks)

Prepared by Dr. Makambi


iv) Model Specification (4 Marks)
v) Strength of the model (2 Marks)
vi) Comment on the regression results (4 Marks)

Question Four

Clearly describe the following STATA commands & the expected output(s)

i) twoway (scatter Y X, mlabel(Y)) (lfit Y X)


ii) twoway(tsline Y in 1/20)
iii) keep if! Missing(X)
iv) ttest Y, by(X) level (99)
v) regress Y X1 X2 i.X3 if X2<30

Prepared by Dr. Makambi

You might also like