Dependent Variable: I
Method: Least Squares
Date: 10/01/16 Time: 08:51
Sample: 1 30
Included observations: 30
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
Y
R
6.224938
0.769911
-0.184196
2.510894
0.071791
0.126416
2.479172
10.72442
-1.457068
0.0197
0.0000
0.1566
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.816282
0.802673
3.329642
299.3358
-77.07369
59.98221
0.000000
20.22200
7.495569
5.338246
5.478366
5.383071
0.852153
10
2
0
-2
-4
-6
-8
2
10
12
Residual
14
16
18
20
22
Actual
24
26
28
30
Fitted
6.764643
10.53429
Prob. F(2,25)
Prob. Chi-Square(2)
0.0045
0.0052
Test Equation:
Dependent Variable: RESID
Method: Least Squares
Date: 10/01/16 Time: 08:56
Sample: 1 30
Included observations: 30
Presample missing value lagged residuals set to zero.
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
Y
1.799886
-0.012173
2.202606
0.060373
0.817162
-0.201637
0.4216
0.8418
R
RESID(-1)
RESID(-2)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
-0.107141
0.697416
-0.165681
0.351143
0.247326
2.787301
194.2262
-70.58555
3.382321
0.024177
0.112469
0.208636
0.204936
-0.952623
3.342738
-0.808452
0.3499
0.0026
0.4265
-2.35E-15
3.212775
5.039036
5.272569
5.113746
1.868584
Teste Breusch-Godfrey
H:0 -> no tem autocorrelao entre os resduos, logo (coeficientes de RESID (-1)
e RESID (-2) conjuntamente so iguais a zero);
H:1 -> tem autocorrelao, pelo menos um dos coeficientes diferente de zero
conjuntamente.
O valor F calculado de 6,76 com p-valor muito pequeno (<<1%), logo, existe
evidencia estatstica de autocorrelao entre os resduos, dessa forma, evidncia
de impacto no modelo.
Como fazer na mo
Rodar modelo, criar srie de resduos.
Roda esse modelo: >>> resid01 c y r resid01(-1) resid01(-2)
Coefficient
Std. Error
t-Statistic
Prob.
C
Y
R
RESID01(-1)
RESID01(-2)
1.820885
-0.002345
-0.127743
0.716243
-0.178517
2.425018
0.067168
0.118703
0.219487
0.212598
0.750875
-0.034910
-1.076159
3.263266
-0.839690
0.4603
0.9725
0.2930
0.0034
0.4097
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.361424
0.250368
2.856312
187.6459
-66.36321
3.254416
0.029668
-0.108571
3.298994
5.097372
5.335266
5.170098
1.879764
Coefficient
Std. Error
t-Statistic
Prob.
C
Y
R
6.224938
0.769911
-0.184196
2.124542
0.077516
0.121236
2.930014
9.932311
-1.519318
0.0068
0.0000
0.1403
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
Prob(Wald F-statistic)
0.816282
0.802673
3.329642
299.3358
-77.07369
59.98221
0.000000
0.000000
20.22200
7.495569
5.338246
5.478366
5.383071
0.852153
57.35421
e) Qual o nvel de investimento do prximo ano dado que R=14 e Y=36? Faa
a projeo com o modelo com e sem ajuste para autocorrelao e
compare os resultados.
I = 6,2249 + 0,76*36 0,1841*14 = X valor na calculadora
12.6)
O arquivo ICECR.WFL possui variveis potencialmente importantes para modelar
a demanda por sorvete. As variveis so:
Q = consumo de sorvete per capita
P = preo do sorvete
I = renda familiar
F: temperatura mdia
Considerando estas informaes:
a) Estime o seguinte modelo:
Qt = 1 + 2Pt + 3It + 4Ft + et
Dependent Variable: Q
Method: Least Squares
Date: 10/01/16 Time: 16:28
Sample: 1 30
Included observations: 30
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
P
I
F
0.197315
-1.044414
0.003308
0.003458
0.270216
0.834357
0.001171
0.000446
0.730212
-1.251759
2.823722
7.762213
0.4718
0.2218
0.0090
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.718994
0.686570
0.036833
0.035273
58.61944
22.17489
0.000000
0.359433
0.065791
-3.641296
-3.454469
-3.581528
1.021170
c) H evidencia de autocorrelao dos resduos? Faa o teste de DurbinWatson e multiplicador de Lagrange. Os valores crticos so: d L=1,214 e
dU=1,650.
O teste de Durbin-Watson no aplicvel nesse exerccio, sendo adotado como proxy o valor do teste BreuschGodfrey.
Breusch-Godfrey Serial Correlation LM Test:
F-statistic
Obs*R-squared
2.110591
4.487249
Prob. F(2,24)
Prob. Chi-Square(2)
0.1431
0.1061
Test Equation:
Dependent Variable: RESID
Method: Least Squares
Date: 10/01/16 Time: 16:31
Sample: 1 30
Included observations: 30
Presample missing value lagged residuals set to zero.
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
P
I
F
RESID(-1)
RESID(-2)
0.048078
-0.164820
2.80E-05
-8.48E-05
0.467604
-0.127675
0.262663
0.805036
0.001164
0.000503
0.229323
0.263176
0.183041
-0.204736
0.024051
-0.168684
2.039059
-0.485130
0.8563
0.8395
0.9810
0.8675
0.0526
0.6320
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.149575
-0.027597
0.035354
0.029997
61.04972
0.844236
0.531897
-1.05E-16
0.034876
-3.669981
-3.389742
-3.580330
1.627763
Coefficient
Std. Error
t-Statistic
Prob.
C
P
I
F
0.197315
-1.044414
0.003308
0.003458
0.337811
0.977249
0.001328
0.000400
0.584099
-1.068728
2.491080
8.640830
0.5642
0.2950
0.0194
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
Prob(Wald F-statistic)
0.718994
0.686570
0.036833
0.035273
58.61944
22.17489
0.000000
0.000000
0.359433
0.065791
-3.641296
-3.454469
-3.581528
1.021170
27.63410