Upload
yolandafitri-zulvia
View
904
Download
1
Embed Size (px)
DESCRIPTION
Citation preview
No Absen 4Nama: Yolandafitri ZulviaNPM: 120420100004Ujian Akhir Ekonometrika Jurusan: Magister Ekonomi ManajemenDosen: Nury EffendiHari/tgl: Rabu 5 Januari 2011
1. Soal Nomor C6.2 buku WoolridgeUse the data in WAGE1.RAW for this exercise.(i) Use OLS to estimate the equation
Log(wage) βo+β1educ+ β2exper+ β3exper2
and report the results using the usual format.
JAWAB:
Dependent Variable: LOG(WAGE)Method: Least SquaresDate: 01/05/11 Time: 09:18Sample: 1 526Included observations: 526
Variable Coefficient Std. Error t-Statistic Prob.
C 0.127998 0.105932 1.208296 0.2275EDUC 0.090366 0.007468 12.10041 0.0000
EXPER 0.041009 0.005197 7.891606 0.0000EXPER^2 -0.000714 0.000116 -6.163888 0.0000
R-squared 0.300273 Mean dependent var 1.623268Adjusted R-squared 0.296251 S.D. dependent var 0.531538S.E. of regression 0.445906 Akaike info criterion 1.230158Sum squared resid 103.7904 Schwarz criterion 1.262594Log likelihood -319.5316 Hannan-Quinn criter. 1.242858F-statistic 74.66829 Durbin-Watson stat 1.785009Prob(F-statistic) 0.000000
Estimation Command:=========================LS LOG(WAGE) C EDUC EXPER EXPER^2
Estimation Equation:=========================LOG(WAGE) = C(1) + C(2)*EDUC + C(3)*EXPER + C(4)*EXPER^2
Substituted Coefficients:=========================LOG(WAGE) = 0.127997507231 + 0.0903658157891*EDUC + 0.041008875312*EXPER - 0.000713558157785*EXPER^2
R-squared 0.300273 N 526
ii) Is exper2 statistically significant at the 1% level?Jawab:Exper2 significant dilihat dari probabilitasnya 0.0000 tapi dia memiliki
koefisien (-) yaitu -0.000714 artinya dia mempunyai pengaruh negative terhadap wage. Kalau exper (+) artinya semakin banyak experience atau pengalaman akan meningkatkan gaji.
iii)
IV) At what value of exper does additional experience actually lower predictedlog(wage)? How many people have more experience in thissample?
2. SOAL C6.5Use the housing price data in HPRICE1.RAW for this exercise.Log(price)= β0+ β1 log(lotsize) + β2 log(sqrft) + β3bdrms + u
Jawab:
Dependent Variable: LOG(PRICE)Method: Least SquaresDate: 01/05/11 Time: 09:48Sample: 1 88Included observations: 88
Variable Coefficient Std. Error t-Statistic Prob.
C -1.297042 0.651284 -1.991516 0.0497LLOTSIZE 0.167967 0.038281 4.387712 0.0000LSQRFT 0.700232 0.092865 7.540305 0.0000BDRMS 0.036958 0.027531 1.342413 0.1831
R-squared 0.642965 Mean dependent var 5.633180Adjusted R-squared 0.630214 S.D. dependent var 0.303573S.E. of regression 0.184603 Akaike info criterion -0.496833Sum squared resid 2.862564 Schwarz criterion -0.384227Log likelihood 25.86065 Hannan-Quinn criter. -0.451467F-statistic 50.42373 Durbin-Watson stat 2.088995Prob(F-statistic) 0.000000
Estimation Command:=========================LS LOG(PRICE) C LLOTSIZE LSQRFT BDRMS
Estimation Equation:=========================LOG(PRICE) = C(1) + C(2)*LLOTSIZE + C(3)*LSQRFT + C(4)*BDRMS
Substituted Coefficients:=========================LOG(PRICE) = -1.29704178525 + 0.167966674526*LLOTSIZE + 0.700232436031*LSQRFT + 0.0369583833496*BDRMS
ii)Find the predicted value of log( price), when lotsize _ 20,000, sqrft _2,500, and bdrms _ 4. Using the methods in Section 6.4, find the predictedvalue of price at the same values of the explanatory variables.
JAWAB: Lotsize = 20000, maka pricenya = 0.167967*ln(20000)= 1.663459
EXP(1.663459)= 5.277535Jadi, lotsize naik 20000 akan meningkatkan price 5.277535
Sqrft= 2500, maka pricenya = 0.700232*ln(2500)= 5.478647 EXP(5.478647)= 239.522
Jadi, sqrft naik 2500 akan meningkatkan price = 239.522 Bdrms= 4, maka pricenya= 0.036958*ln(4)= 0.051235
EXP(0.051235)=1.05257Jadi, bdrms naik 4 akan meningkatkan price =1.05257
iii). For explaining variation in price, decide whether you prefer the model from part (i) or the modelprice= β0+ β1 lotsize + β2 sqrft + β3bdrms + u
Dependent Variable: PRICEMethod: Least SquaresDate: 01/05/11 Time: 10:11Sample: 1 88Included observations: 88
Variable Coefficient Std. Error t-Statistic Prob.
C -21.77031 29.47504 -0.738601 0.4622LOTSIZE 0.002068 0.000642 3.220096 0.0018SQRFT 0.122778 0.013237 9.275093 0.0000BDRMS 13.85252 9.010145 1.537436 0.1279
R-squared 0.672362 Mean dependent var 293.5460Adjusted R-squared 0.660661 S.D. dependent var 102.7134S.E. of regression 59.83348 Akaike info criterion 11.06540Sum squared resid 300723.8 Schwarz criterion 11.17800Log likelihood -482.8775 Hannan-Quinn criter. 11.11076F-statistic 57.46023 Durbin-Watson stat 2.109796Prob(F-statistic) 0.000000
Estimation Command:=========================LS PRICE C LOTSIZE SQRFT BDRMS
Estimation Equation:=========================PRICE = C(1) + C(2)*LOTSIZE + C(3)*SQRFT + C(4)*BDRMS
Substituted Coefficients:=========================PRICE = -21.7703086036 + 0.00206770660199*LOTSIZE + 0.122778185222*SQRFT + 13.8525218631*BDRMS
3.SOAL C7.2(i) Estimate the modelLog (wage)= β + β1 educ + β2 exper + β3 tenure + β4 married + β5 black + β6 south + β7 urban + uand report the results in the usual form. Holding other factors fixed, whatis the approximate difference in monthly salary between blacks and nonblacks?Is this difference statistically significant?JAWAB
Dependent Variable: LOG(WAGE)Method: Least SquaresDate: 01/05/11 Time: 10:20Sample: 1 935Included observations: 935
Variable Coefficient Std. Error t-Statistic Prob.
C 5.395497 0.113225 47.65286 0.0000EDUC 0.065431 0.006250 10.46826 0.0000
EXPER 0.014043 0.003185 4.408852 0.0000TENURE 0.011747 0.002453 4.788998 0.0000MARRIED 0.199417 0.039050 5.106691 0.0000
BLACK -0.188350 0.037667 -5.000444 0.0000SOUTH -0.090904 0.026249 -3.463193 0.0006URBAN 0.183912 0.026958 6.822087 0.0000
R-squared 0.252558 Mean dependent var 6.779004Adjusted R-squared 0.246914 S.D. dependent var 0.421144S.E. of regression 0.365471 Akaike info criterion 0.833260Sum squared resid 123.8185 Schwarz criterion 0.874676Log likelihood -381.5490 Hannan-Quinn criter. 0.849052F-statistic 44.74706 Durbin-Watson stat 1.822637Prob(F-statistic) 0.000000
(i) Sig signifikan semuanya Black orang hitam digaji 19% lebih rendah dibandingkan dengan orang kulit lain non black (putih)Selatan orang selatan digaji 9% lebih rendah
(ii) Add the variables exper2 and tenure2 to the equation and show that theyare jointly insignificant at even the 20% level.
JAWAB:
Dependent Variable: LOG(WAGE)Method: Least SquaresDate: 01/05/11 Time: 10:25Sample: 1 935Included observations: 935
Variable Coefficient Std. Error t-Statistic Prob.
C 5.358676 0.125914 42.55812 0.0000EDUC 0.064276 0.006311 10.18400 0.0000
EXPER 0.017215 0.012614 1.364747 0.1727TENURE 0.024929 0.008130 3.066433 0.0022MARRIED 0.198547 0.039110 5.076585 0.0000
BLACK -0.190664 0.037701 -5.057240 0.0000SOUTH -0.091215 0.026236 -3.476774 0.0005URBAN 0.185424 0.026959 6.878122 0.0000
EXPER^2 -0.000114 0.000532 -0.213964 0.8306TENURE^2 -0.000796 0.000471 -1.690923 0.0912
R-squared 0.254958 Mean dependent var 6.779004Adjusted R-squared 0.247709 S.D. dependent var 0.421144S.E. of regression 0.365278 Akaike info criterion 0.834322Sum squared resid 123.4210 Schwarz criterion 0.886092Log likelihood -380.0455 Hannan-Quinn criter. 0.854062F-statistic 35.17112 Durbin-Watson stat 1.819339Prob(F-statistic) 0.000000
Exper (+) semakin banyak experience semakin meningkat gajinya.Koefisien pada tenure2
(-)Koefisien pada exper2
(-)
(iii) Extend the original model to allow the return to education todepend on race and test whether the return to education does dependon race.JAWAB
Dependent Variable: LOG(WAGE)Method: Least SquaresDate: 01/05/11 Time: 10:34Sample: 1 935Included observations: 935
Variable Coefficient Std. Error t-Statistic Prob.
C 5.374817 0.114703 46.85866 0.0000EDUC 0.067115 0.006428 10.44160 0.0000
EXPER 0.013826 0.003191 4.333276 0.0000TENURE 0.011787 0.002453 4.805362 0.0000MARRIED 0.198908 0.039047 5.094007 0.0000
BLACK 0.094809 0.255399 0.371217 0.7106SOUTH -0.089450 0.026277 -3.404111 0.0007URBAN 0.183852 0.026955 6.820800 0.0000
EDUC*BLACK -0.022624 0.020183 -1.120943 0.2626
R-squared 0.253571 Mean dependent var 6.779004Adjusted R-squared 0.247122 S.D. dependent var 0.421144S.E. of regression 0.365420 Akaike info criterion 0.834043Sum squared resid 123.6507 Schwarz criterion 0.880636Log likelihood -380.9150 Hannan-Quinn criter. 0.851809F-statistic 39.32158 Durbin-Watson stat 1.826713Prob(F-statistic) 0.000000
Estimation Command:=========================LS LOG(WAGE) C EDUC EXPER TENURE MARRIED BLACK SOUTH URBAN EDUC*BLACK
Estimation Equation:=========================LOG(WAGE) = C(1) + C(2)*EDUC + C(3)*EXPER + C(4)*TENURE + C(5)*MARRIED + C(6)*BLACK + C(7)*SOUTH + C(8)*URBAN + C(9)*EDUC*BLACK
Substituted Coefficients:=========================LOG(WAGE) = 5.37481703029 + 0.0671153308552*EDUC + 0.0138258814311*EXPER + 0.0117870227642*TENURE + 0.198907694212*MARRIED + 0.0948086755005*BLACK - 0.0894495435054*SOUTH + 0.183852289256*URBAN - 0.0226236090572*EDUC*BLACK
Black dan educ tidak signifikan. Artinya Pendidikan akan meningkatkan gaji tidak melihat warna kulitanya apakah hitam, latin atau asia.
(IV) Again, start with the original model, but now allow wages to differacross four groups of people: married and black, married and nonblack,single and black, and single and nonblack. What is the estimated wagedifferential between married blacks and married nonblacks?
JAWAB
Dependent Variable: LOG(WAGE)Method: Least SquaresDate: 01/05/11 Time: 10:38Sample: 1 935Included observations: 935
Variable Coefficient Std. Error t-Statistic Prob.
C 5.403793 0.114122 47.35093 0.0000EDUC 0.065475 0.006253 10.47095 0.0000
EXPER 0.014146 0.003191 4.433117 0.0000TENURE 0.011663 0.002458 4.744941 0.0000MARRIED 0.188915 0.042878 4.405892 0.0000
BLACK -0.240820 0.096023 -2.507943 0.0123SOUTH -0.091989 0.026321 -3.494879 0.0005URBAN 0.184350 0.026978 6.833394 0.0000
MARRIED*BLACK 0.061354 0.103275 0.594083 0.5526
R-squared 0.252842 Mean dependent var 6.779004Adjusted R-squared 0.246388 S.D. dependent var 0.421144S.E. of regression 0.365599 Akaike info criterion 0.835018Sum squared resid 123.7714 Schwarz criterion 0.881611Log likelihood -381.3708 Hannan-Quinn criter. 0.852784F-statistic 39.17047 Durbin-Watson stat 1.824148Prob(F-statistic) 0.000000
Estimation Command:=========================LS LOG(WAGE) C EDUC EXPER TENURE MARRIED BLACK SOUTH URBAN MARRIED*BLACK
Estimation Equation:=========================LOG(WAGE) = C(1) + C(2)*EDUC + C(3)*EXPER + C(4)*TENURE + C(5)*MARRIED + C(6)*BLACK + C(7)*SOUTH + C(8)*URBAN + C(9)*MARRIED*BLACK
Substituted Coefficients:=========================LOG(WAGE) = 5.40379326745 + 0.065475113325*EDUC + 0.0141462059065*EXPER + 0.0116628070316*TENURE + 0.188914701141*MARRIED - 0.240819977672*BLACK - 0.0919894174516*SOUTH + 0.184350063352*URBAN + 0.0613536984779*MARRIED*BLACK
Jadi married*black tidak signifikan artinya status pernikahan akan meningkatkan gaji tidak melihat warna kulitnya.