Download Final Exam, with Answer Key - Statistics for Engineering | STAT 4706 and more Exams Statistics in PDF only on Docsity!
<< f
Stat 4706 Name
Exam 1
Spring 2000
Show all work on these pages for full credit.
You are allowed to use a calculator, the tables provided, and a
formula sheet (one 8 1/2 x 11 sheet, both sides, formulas only.)
Otherwise, this is a closed book exam.
1. An industrial engineer wishes to establish the relationship between
y = cost per production of a batch of laminated wafers, in dollars
x = size of the batch, measured in number of wafers
For n= 20 batches, the following calculations resulted:
Sux = 75,710.0 Syy = 5,946,702.0 Sxy = 659,553.0
Ya = 3,340.0 >» = 37,130.0
(a) Give the Least Squares regression equation.
6595532, _ ga
Ai = WIE
A= Ue @ 0.91 (22GB) = 856.5 - B.21(6Y
oe
2 56S - HSE
= 01.67
A
Ya ¥Ol b7 thi E
(b) Give the regression variance, s?
2 (454553) xi 4553)
t Jeg,
A, = F956 18 - —aeaee 7570
L [stal7e> -$7¢8, 740-4]
gle ‘ 7 ,
£009 EF b
48
= MME
3. Continued.
. (a) Based on the experimental data, give the regression function that
explains the cost of this software development.
A
Ye GH t G63 Ey tO18 Ky +0, 20k
(b) Do the 3 regressor variables collectively have a significant effect of
explaining the cost of this software development? Justify your answer.
fF xltbi3on wt a P- walee of 0.0607 wheck
(c) Give the regression variance, 5? :
2
as = 346.3
(d) If xz is the last variable to be entered in the regression model, would
it have a significant effect in explaining the cost of developing this
software? Justify your answer.
to, Me fiat of hipre tao 2,707
wih 2 Fe wate Ff 8,4837 wndeiatecy 1. y
mt enrhenee. eT Va far wn feed am
syplasing Te cont of dewclaping tin. afftone
phan 7, wok ty tae, sally x fa mets
4, The engineering consulting firm (Problem 3) felt that an improvement of
the regression model could be made by the inclusion of other regressors.
Two additional (secret) variables were included and a stepwise selection
procedure was run to see if a "better" model could be obtained. The SAS
output of this procedure is:
Cost of Software Development
Stepwise Procedure for Dependent variable ¥
Step 1 Varisble x3 Entered
R-aquare = 0.82478630
C(p) = 18.20077678
OF Sum of Squares Mean Square F Prob r
Regreasion a 15705,57436542 15705.57436542 51.78 06,0003
Error ia 3336.42794227 303.31072202
Total 12 19041. 99230769
Standard Type II
Varieble Error Sum of Squares F ProboF
IWTERCEP 12.40428457 6. 96912667 960.69180139 3.17 0.1027
x3 2.01281099 0,44074902 © 15705.57436542
Step 2 Variable x4 Entered
R-mquare = 0.90598356
51.78 0.0001
c(p) = 7.59543473
oF Sum of Squares Mean Square F ProboF
Regreasion 2 17251.73202617 —-8625..26601309 48.18 9.0002
Error 10 1790,26028152 179.02602815
total az 19042.99230769
Parmmster Standard ‘Type II
Variable Eatimate Error Sum of Squares F ProbF
INTERCEP -17.99417635 11.64743361 427 29695) 2.39 0.1534
x3 2.75265331 0.60182066 = -3745.20637482 20.92 0.0010
xa -0.73367784 0,24365289 = -1566.15766075 a.64 0.0148
Step 2 Variable X5 Entered
R-square = 0,92541601
Clip) = 6.57867465
or Sum of Squares Mean Square F Proper
Regression 3 17621.76436345 9 5873,92152115 37.22 0.0002
Error 2 1420.22774424 157.80308269
Total 12 19041.99230765
Parameter Standard ‘ype IT
Variable Estimate Error sun of Squares YF Prob>F
INTERCEP -25.56821922 12.00188288 716.17816153 4.54 0.0620
x 3.42038073 0.71371711 = -3624.20008550 22.97 09,0010
m +0.93410669 0.26845760 © 1910.54398281 12.11 0.0069
Comt of Software Development
2
5 =0.01774399 0,01150745 270.03253728 2.34 0.1602
R-square = 0,90598356
Cp) = 7.59543473
or Sum of Squares Mean Square F Prob>r
Regression a 47251.73202617 862586601309 48.28 0.0002
Error 20 1790.26028152 179.02602815
Fotal 12 19041.99230769
4. Continued
i Pacamater Standard
Variable Estimate Error ¥ Probr
INTERCEP -27,.99417635 11.64743361 2.39) 0.1534
x 2.75265331 0, 69182066 20.92 0.9020
Fey -0,73367784 0.24965289 «= -1546.15766075 9.64 0,0148
| All variables left in the model are significant at the 0.1500 level.
The stepwise mathod terminated because the next variable to be entered was just removed.
Sumary of Stepwiee Procedure for Dependent Verieble ¥
variable member Partial = Model
Step Entered Removed In RD Rea tp) F OProb>F
1 x3 1 «0.8248 «0.9248 «18.2008 «= 51.7805 9.0001
| 2 x 2 0,0812 0.9060 = 7.5954 8.6365 0.0148
3 25 30,0194 «0.9254 = 6.8787 2.3449 0.1601
‘ aS 2 0.0194 0.9060 7, 5954 2.3449 0.1602
(a) Give the final model obtained by the stepwise procedure.
He LIDGI PE BAF Beg — OTF Eye
What is the R? value for this model?
2
i Rae, Yee
What is the regression variance for this final model?
! AL 247%03
(b) If possible, give the simplest model that has an R? value of at least 0.90.
: A
-ITGG + A.IS Vg ~ F073 Key