1. Buka data → File → Open → Foreign Data as workfile → Pilih Datanya → Next → Finish
2. Pilih Menu Proc → Stucture/Resize Current Page
3. Pada worfile structure type: klik dropdown pilih dated panel
Pada frequency: klik dropdown sesuai frekuensi data kita, misalkan annual (tahunan)
Pada cross section ID series: ketik nama judul kolom cross section missal “”;
Pada date series : ketik nama judul kolom time series missal “tahun”
Klik Ok.
4. Untuk persamaan model common/ pooled effect
Menu Quick → Estimate → equation
Pilih specification
Pada kotak Equation specification
Ketik persamaan panel yang akan kita buat: np c uk k_aud k_ind
Pada method: LS-Least Squares
Klik OK
Akan Akan muncul hasil sebagai berikut:
|
Dependent Variable: NP
|
||||
|
Method: Panel Least Squares
|
||||
|
Date: 02/01/14 Time: 11:28
|
||||
|
Sample: 2008 2012
|
||||
|
Periods included: 5
|
||||
|
Cross-sections included: 5
|
||||
|
Total panel (balanced) observations: 25
|
||||
|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
Prob.
|
|
C
|
-6.134655
|
3.378708
|
-1.815680
|
0.0837
|
|
UK
|
0.094687
|
0.184124
|
0.514256
|
0.6124
|
|
K_AUD
|
2.808153
|
2.214165
|
1.268267
|
0.2186
|
|
K_IND
|
12.93182
|
3.883847
|
3.329643
|
0.0032
|
|
R-squared
|
0.347868
|
Mean dependent var
|
1.573600
|
|
|
Adjusted R-squared
|
0.254706
|
S.D. dependent var
|
1.480484
|
|
|
S.E. of regression
|
1.278108
|
Akaike info criterion
|
3.474285
|
|
|
Sum squared resid
|
34.30474
|
Schwarz criterion
|
3.669305
|
|
|
Log likelihood
|
-39.42856
|
Hannan-Quinn criter.
|
3.528375
|
|
|
F-statistic
|
3.734021
|
Durbin-Watson stat
|
0.611647
|
|
|
Prob(F-statistic)
|
0.026995
|
|||
Untuk persamaan model fixed effect
Menu Quick → Estimate equation
Pilih Panel Option
Pada effect specification:
Cross section: pilih fixed
Pada periode: pilih none
Pada weight:pilih no weight
Pada covariance method: pilih ordinary
Klik OK
Akan muncul hasil sebagi berikut:
|
Dependent Variable: NP
|
||||
|
Method: Panel Least Squares
|
||||
|
Date: 02/01/14 Time: 10:53
|
||||
|
Sample: 2008 2012
|
||||
|
Periods included: 5
|
||||
|
Cross-sections included: 5
|
||||
|
Total panel (balanced) observations: 25
|
||||
|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
Prob.
|
|
C
|
-4.529566
|
14.07672
|
-0.321777
|
0.7515
|
|
UK
|
0.744648
|
1.517746
|
0.490627
|
0.6300
|
|
K_AUD
|
1.877844
|
4.149925
|
0.452501
|
0.6566
|
|
K_IND
|
2.096274
|
16.37727
|
0.127999
|
0.8997
|
|
Effects Specification
|
||||
|
Cross-section fixed (dummy variables)
|
||||
|
R-squared
|
0.535274
|
Mean dependent var
|
1.573600
|
|
|
Adjusted R-squared
|
0.343916
|
S.D. dependent var
|
1.480484
|
|
|
S.E. of regression
|
1.199178
|
Akaike info criterion
|
3.455487
|
|
|
Sum squared resid
|
24.44646
|
Schwarz criterion
|
3.845527
|
|
|
Log likelihood
|
-35.19358
|
Hannan-Quinn criter.
|
3.563667
|
|
|
F-statistic
|
2.797237
|
Durbin-Watson stat
|
0.671849
|
|
|
Prob(F-statistic)
|
0.039374
|
|||
Untuk melihat efek cross section nilai konstanta per perusahaan dapat dilihat dengan cara:
Klik menu view → Fixed/Randon effect → Cross section Effect
Untuk persamaan model random effect
Klik Menu Quick → Estimate equation
Pilih Panel Option
Pada effect specification:
Cross section: pilih random
Pada periode: pilih random
Pada weight:pilih no weight
Pada covariance method: pilih ordinary
Klik OK
Akan muncul hasil sebagi berikut:
|
Dependent Variable: NP
|
||||
|
Method: Panel EGLS (Two-way random effects)
|
||||
|
Date: 02/01/14 Time: 20:30
|
||||
|
Sample: 2008 2012
|
||||
|
Periods included: 5
|
||||
|
Cross-sections included: 5
|
||||
|
Total panel (balanced) observations: 25
|
||||
|
Swamy and Arora estimator of component variances
|
||||
|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
Prob.
|
|
C
|
-4.937867
|
3.756465
|
-1.314498
|
0.2029
|
|
UK
|
0.076919
|
0.212785
|
0.361485
|
0.7213
|
|
K_AUD
|
2.227142
|
2.476596
|
0.899276
|
0.3787
|
|
K_IND
|
11.23748
|
4.354077
|
2.580910
|
0.0174
|
|
Effects Specification
|
||||
|
S.D.
|
Rho
|
|||
|
Cross-section random
|
0.388346
|
0.1415
|
||
|
Period random
|
0.259552
|
0.0632
|
||
|
Idiosyncratic random
|
0.920841
|
0.7954
|
||
|
Weighted Statistics
|
||||
|
R-squared
|
0.242540
|
Mean dependent var
|
1.040656
|
|
|
Adjusted R-squared
|
0.134332
|
S.D. dependent var
|
1.225853
|
|
|
S.E. of regression
|
1.140549
|
Sum squared resid
|
27.31791
|
|
|
F-statistic
|
2.241413
|
Durbin-Watson stat
|
0.635036
|
|
|
Prob(F-statistic)
|
0.113253
|
|||
|
Unweighted Statistics
|
||||
|
R-squared
|
0.341509
|
Mean dependent var
|
1.573600
|
|
|
Sum squared resid
|
34.63922
|
Durbin-Watson stat
|
0.567315
|
|
Untuk mengetahui efek cross section nilai konstanta per perusahaan dapat dilihat dengan cara:
Klik menu view-->Fixed/Randon effect-->Period Effect
|
2008-01-01
|
-0.322352
|
|
2009-01-01
|
-0.208940
|
|
2010-01-01
|
0.024149
|
|
2011-01-01
|
0.195685
|
|
2012-01-01
|
0.311459
|
Untuk menguji antara model fix dan ramdom dignakan uji Hausman Test
Carannya: klik fix/random effect testing → Redundant Fix Effect Likelihood Ratio
Terlihat nilainya tidak signifika sehingga random yang terpilih:
|
Redundant Fixed Effects Tests
|
||||
|
Equation: Untitled
|
||||
|
Test cross-section fixed effects
|
||||
|
Effects Test
|
Statistic
|
d.f.
|
Prob.
|
|
|
Cross-section F
|
1.713856
|
(4,17)
|
0.1933
|
|
|
Cross-section Chi-square
|
8.469959
|
4
|
0.0758
|
|
Untuk menguji antara model pooled/common dan ramdom dignakan uji Hausman Test
Caranya: klik fix/random effect testing correlated Random Effect-Hausman Test
|
Correlated Random Effects - Hausman Test
|
||||
|
Equation: Untitled
|
||||
|
Test cross-section and period random effects
|
||||
|
Test Summary
|
Chi-Sq. Statistic
|
Chi-Sq. d.f.
|
Prob.
|
|
|
Cross-section random
|
8.939450
|
3
|
0.0301
|
|
|
Period random
|
0.000000
|
3
|
1.0000
|
|
|
Cross-section and period random
|
3.096451
|
3
|
0.3770
|
|
|
* Period test variance is invalid. Hausman statistic set to zero.
|
||||
Terlihat bahwa model pooled/common yang terpilih
Sehingga yang akan diuji adalah model pooled/common effect
Untuk uji normalitas dipilih view → residual test → Histogram-Normality Test
0 komentar:
Posting Komentar