Springer Texts in Business and Economics
Dependent Variable: LNC
 |
 |
 |
Analysis of Variance
|
Sum of |
Mean |
||||
|
Source |
DF |
Squares |
Square |
F Value |
Prob > F |
|
Model |
1 |
0.04693 |
0.04693 |
1.288 |
0.2625 |
|
Error |
44 |
1.60260 |
0.03642 |
||
|
C Total |
45 |
1.64953 |
|||
|
Root MSE |
0.19085 |
R-square |
0.0285 |
||
|
Dep Mean |
4.84784 |
Adj R-sq |
0.0064 |
||
|
C. V. |
3.93675 |
|
Parameter Estimates |
INTERCEP 1 5.931889 0.95542530
LNY 1 -0.227003 0.19998321
The income elasticity is —0.227 which is negative! Its standard error is (0.1999) and the t-statistic fortesting this income elasticity is zero is —1.135 which is insignificant with a p-value of 0.26. Hence, we cannot reject the null hypothesis. R2 = 0.0285 and s = 0.19085. This regression is not very useful. The income variable is not significant and the R2 indicates that the
regression explains only 2.8% of the variation in consumption. b. Plot of Residuals, and the 95% confidence interval for the predicted value.
SAS Program for 3.14 Data CIGARETT;
Input OBS STATE $ LNC LNP LNY;
Cards;
Proc reg data=cigarett;
model lnc=lny;
*plot residual. *lny=’*’;
*plot (U95. L95.)*lny=’-’ p.*lny/overlay symbol=‘*’;
output out=out1 r=resid p=pred u95=upper95 195=lower95;
proc plot data=out1 vpercent=75 hpercent=100; plot resid*lny=‘*’;
proc plot data=out1 vpercent=95 hpercent=100; plot (Upper95 Lower95)*lny=‘-’ Pred*lny=‘*’
/overlay;
run;
