You are on page 1of 6

The data are given in the file <Data_HW4.

xlsx>

This data set of size n = 17 contains observations from a small assembly plant. The predictors
are
x1=average monthly temperature of the plant (degrees Fahrenheit),
x2=amount of production (in pounds),fmn and the response is
Y =the monthly water usage (in gallons).

(1) Perform the stepwise regression analysis by hand, that is, following Slides 20-30 in
Chapter 5.
In your stepwise regression, set
Alpha-to-Enter: 0.15 ; Alpha-to-Remove: 0.30
In each step, you need to clearly state if a variable has been entered or removed,
and explain
Y why a variable has been entered/removed, or
Y why no variable qualifies to enter or leave.

Step 1:
Based on (1.1), (1.2), (1.3), (1.4), (1.5),
x2 is chosen to enter the model, since
it has the highest t value, 9.91 (corresponding to the lowest p value);
the p value is 0.007, less than the Alpha to Enter 0.15.

Step 2:
Based on (2.1), (2.2), (2.3), (2.4),
x4 is chosen to enter the model since
it has the highest t value 9.44 (corresponding to the lowest p value);
the p value is 0.003, less than the Alpha to Enter 0.15.
No variable is removed in this step, since
In the model with x2 and x4, x2 is still significant with a high t value 3.64,
which corresponds to a low p value of 0.003. The p value is less than Alpha to
Remove 0.15.

1
(2) Suppose that you have performed the best subset regression analysis on the dataset, and
obtained the following outputs from MINITAB:

Compare the 9 candidate models based on the four criteria:


R-Sq, R-Sq(adj), Cp, s
You can refer to Slides 37-39, Chapter 5.

2
APPENDIX

(1.1) The regression equation is Y = 2691 + 9.45 X1

Predictor Coef SE Coef T P


Constant 2691.0 541.3 4.97 0.000
X1 9.448 8.181 1.15 0.266

S = 442.111 R-Sq = 8.2% R-Sq(adj) = 2.0%

Analysis of Variance

Source DF SS MS F P
Regression 1 260702 260702 1.33 0.266
Residual Error 15 2931930 195462
Total 16 3192632

(1.2) The regression equation is Y = 2273 + 0.0799 X2

Predictor Coef SE Coef T P


Constant 2273.1 338.7 6.71 0.000
X2 0.07989 0.02538 3.15 0.007

S = 358.000 R-Sq = 39.8% R-Sq(adj) = 35.8%

Analysis of Variance

Source DF SS MS F P
Regression 1 1270172 1270172 9.91 0.007
Residual Error 15 1922459 128164
Total 16 3192632

(1.3) The regression equation is Y = 3886 - 27.1 X3

Predictor Coef SE Coef T P


Constant 3886 1690 2.30 0.036
X3 -27.13 78.54 -0.35 0.735

S = 459.525 R-Sq = 0.8% R-Sq(adj) = 0.0%

Analysis of Variance

Source DF SS MS F P
Regression 1 25190 25190 0.12 0.735
Residual Error 15 3167442 211163
Total 16 3192632

(1.4) The regression equation is Y = 1778 + 8.39 X4

Predictor Coef SE Coef T P


Constant 1777.7 874.2 2.03 0.060
X4 8.393 4.775 1.76 0.099

S = 420.113 R-Sq = 17.1% R-Sq(adj) = 11.5%

Analysis of Variance

Source DF SS MS F P
Regression 1 545213 545213 3.09 0.099
Residual Error 15 2647418 176495
Total 16 3192632

3
(1.5) The regression equation is Y = 3352 - 1.07 X5

Predictor Coef SE Coef T P


Constant 3352.2 220.8 15.18 0.000
X5 -1.067 4.188 -0.25 0.802

S = 460.354 R-Sq = 0.4% R-Sq(adj) = 0.0%

Analysis of Variance

Source DF SS MS F P
Regression 1 13749 13749 0.06 0.802
Residual Error 15 3178883 211926
Total 16 3192632

(2.1) The regression equation is Y = 1615 + 0.0808 X2 + 9.96 X1

Predictor Coef SE Coef T P


Constant 1615.5 528.0 3.06 0.008
X2 0.08081 0.02422 3.34 0.005
X1 9.957 6.322 1.57 0.138

S = 341.541 R-Sq = 48.8% R-Sq(adj) = 41.5%

Analysis of Variance

Source DF SS MS F P
Regression 2 1559525 779763 6.68 0.009
Residual Error 14 1633106 116650
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X1 1 289353

(2.2) The regression equation is Y = 3277 + 0.0820 X2 - 48.0 X3

Predictor Coef SE Coef T P


Constant 3277 1349 2.43 0.029
X2 0.08200 0.02587 3.17 0.007
X3 -48.03 62.38 -0.77 0.454

S = 362.961 R-Sq = 42.2% R-Sq(adj) = 34.0%

Analysis of Variance

Source DF SS MS F P
Regression 2 1348259 674129 5.12 0.021
Residual Error 14 1844373 131741
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X3 1 78087

4
(2.3) The regression equation is Y = 4601 + 0.203 X2 - 21.6 X4

Predictor Coef SE Coef T P


Constant 4601 1011 4.55 0.000
X2 0.20343 0.05585 3.64 0.003
X4 -21.567 8.956 -2.41 0.030

S = 311.604 R-Sq = 57.4% R-Sq(adj) = 51.3%

Analysis of Variance

Source DF SS MS F P
Regression 2 1833271 916635 9.44 0.003
Residual Error 14 1359361 97097
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X4 1 563098

(2.4) The regression equation is Y = 2269 + 0.0800 X2 + 0.08 X5

Predictor Coef SE Coef T P


Constant 2268.7 399.8 5.67 0.000
X2 0.07996 0.02643 3.02 0.009
X5 0.077 3.393 0.02 0.982

S = 370.558 R-Sq = 39.8% R-Sq(adj) = 31.2%

Analysis of Variance

Source DF SS MS F P
Regression 2 1270243 635121 4.63 0.029
Residual Error 14 1922389 137313
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X5 1 71

(3.1) The regression equation is Y = 3866 + 0.193 X2 - 19.7 X4 + 8.04 X1

Predictor Coef SE Coef T P


Constant 3866 1103 3.51 0.004
X2 0.19334 0.05435 3.56 0.004
X4 -19.676 8.742 -2.25 0.042
X1 8.036 5.630 1.43 0.177

S = 300.665 R-Sq = 63.2% R-Sq(adj) = 54.7%

Analysis of Variance

Source DF SS MS F P
Regression 3 2017440 672480 7.44 0.004
Residual Error 13 1175191 90399
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X4 1 563098
X1 1 184170

5
(3.2) The regression equation is Y = 6307 + 0.218 X2 - 23.5 X4 - 71.4 X3

Predictor Coef SE Coef T P


Constant 6307 1598 3.95 0.002
X2 0.21790 0.05531 3.94 0.002
X4 -23.547 8.824 -2.67 0.019
X3 -71.39 52.77 -1.35 0.199

S = 302.757 R-Sq = 62.7% R-Sq(adj) = 54.1%

Analysis of Variance

Source DF SS MS F P
Regression 3 2001031 667010 7.28 0.004
Residual Error 13 1191601 91662
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X4 1 563098
X3 1 167760

(3.3) The regression equation is Y = 4699 + 0.205 X2 - 22.0 X4 - 0.93 X5

Predictor Coef SE Coef T P


Constant 4699 1091 4.31 0.001
X2 0.20504 0.05797 3.54 0.004
X4 -21.989 9.357 -2.35 0.035
X5 -0.932 2.981 -0.31 0.759

S = 322.157 R-Sq = 57.7% R-Sq(adj) = 48.0%

Analysis of Variance

Source DF SS MS F P
Regression 3 1843427 614476 5.92 0.009
Residual Error 13 1349205 103785
Total 16 3192632

Source DF Seq SS
X2 1 1270172
X4 1 563098
X5 1 10156

You might also like