2.

Descriptive Statistics
Std.
Mean Deviation N
fatrate 3.572 .8576 50
jantemp 23.9800 12.70785 50

Correlations
fatrate jantemp
Pearson fatrate 1.000 .168
Correlation jantemp .168 1.000
Sig. (1-tailed) fatrate . .121
jantemp .121 .
N fatrate 50 50
jantemp 50 50

Variables Entered/Removeda
Variables Variables
Model Entered Removed Method
1 jantemp b
. Enter
a. Dependent Variable: fatrate
b. All requested variables entered.

Model Summary
Std. Error Change Statistics
Mod R Adjusted of the R Square F Sig. F
el R Square R Square Estimate Change Change df1 df2 Change
1 .168 a
.028 .008 .8541 .028 1.402 1 48 .242
a. Predictors: (Constant), jantemp

ANOVAa
Model Sum of Squares df Mean Square F Sig.
1 Regression 1.023 1 1.023 1.402 .242b
Residual 35.018 48 .730
Total 36.041 49
 a. Dependent Variable: fatrate
b. Predictors: (Constant), jantemp
Coefficientsa
Standard
ized
Unstandardized Coefficie 95.0% Confidence
Coefficients nts Interval for B Correlations
Std. Lower Upper Zero- Partia
Model B Error Beta t Sig. Bound Bound order l Part
1 (Const 3.299 .260 12.68 .000 2.777 3.822
ant) 9
jantem .011 .010 .168 1.184 .242 -.008 .031 .168 .168 .168
p
a. Dependent Variable: fatrate

3.

Descriptive Statistics
Std.
Mean Deviation N
fatrate 3.572 .8576 50
jantemp 23.9800 12.70785 50
density 144.3200 218.83014 50
gasuse 2.478 2.4370 50

Correlations
fatrate jantemp density gasuse
Pearson fatrate 1.000 .168 -.605 -.213
Correlation jantemp .168 1.000 .072 .352
density -.605 .072 1.000 .130
gasuse -.213 .352 .130 1.000
Sig. (1-tailed) fatrate . .121 .000 .069
jantemp .121 . .310 .006
density .000 .310 . .184
gasuse .069 .006 .184 .
N fatrate 50 50 50 50
jantemp 50 50 50 50
density 50 50 50 50
gasuse 50 50 50 50

Variables Entered/Removeda
Variables Variables
Model Entered Removed Method
1 gasuse, . Enter
density,
jantempb
a. Dependent Variable: fatrate
b. All requested variables entered.

Model Summary
Std. Error Change Statistics
Mod R Adjusted of the R Square F Sig. F
el R Square R Square Estimate Change Change df1 df2 Change
1 .679 a
.461 .426 .6497 .461 13.123 3 46 .000
a. Predictors: (Constant), gasuse, density, jantemp

ANOVAa
Sum of
Model Squares df Mean Square F Sig.
1 Regression 16.621 3 5.540 13.123 .000b
Residual 19.420 46 .422
Total 36.041 49
a. Dependent Variable: fatrate
b. Predictors: (Constant), gasuse, density, jantemp
 Coefficientsa
Standard
ized
Unstandardized Coeffici 95.0% Confidence
Coefficients ents Interval for B Correlations
Std. Lower Upper Zero- Partia
Model B Error Beta t Sig. Bound Bound order l Part
1 (Const 3.639 .206 17.67 .000 3.225 4.053
ant) 8
jantem .020 .008 .296 2.557 .014 .004 .036 .168 .353 .277
p
densit -.002 .000 -.595 - .000 -.003 -.001 -.605 -.627 -.590
y 5.452
gasuse -.084 .041 -.240 - .045 -.167 -.002 -.213 -.290 -.223
2.059
a. Dependent Variable: fatrate

Questions
1. SSE(C) =36.041 PC = 3
SSE(A) = 19.420 PA = 4
SSR = 16.621 PA-PC = 1
N-PA =46 PRE =0.461 F = 13.123

2. PREcritical=0.238
Fcritical=2.807

3. Since PRE 0.238<0.461, we failed to reject null hypothesis

Since F 2.807<13.123, we failed to reject null hypothesis

4. ya=3.637+0.02 x 1−0.002 x 2−0.084 x 3

5.

Collinearity Diagnosticsa
Condition Variance Proportions
Model Dimension Eigenvalue Index (Constant) jantemp density gasuse
1 1 2.974 1.000 .02 .02 .04 .04
2 .599 2.229 .01 .02 .93 .06
3 .316 3.069 .14 .07 .01 .88
4 .111 5.170 .84 .89 .01 .03
a. Dependent Variable: fatrate

Since only the gasuse is greater than 5, so we can conclude that there are no inter-related
variables in the multicollinearity.
6. Jantemp will be more useful because it shows positive correlation compare to the other 2
variables which shows negative correlations
7. Jantemp
8. Null hypothesis: β1=β2=β3=0
Alternative hypothesis: Some β is not zero
Test statistics: In simple linear regression, R square value is 0.028 and p=0.242
While in multiple linear regressions, R square value is 0.461 and p<0.001
Conclusion: Since p<0.001, means that in multiple linear regressions the density, gasuse and
jantemp is more significant predictor for fatrate, we reject the null hypothesis that not every β is
zero.

9. There is a change in R square in simple linear regression and multiple linear regression, but
Jantemp will still affect the fatrate.