Scribd
Upload
111 views28 pages

6.1 A. Estimate The Factor Effects. Which Effects Appear To Be Large?

The document reports the results of a statistical analysis with factors A, B, C, and D. Factors A, C, D, AB, AD, ABC, and ABD were found to have a significant effect on the response variable, while factors B, AC, BC, BD, CD, ACD, and BCD did not have a significant effect. The analysis included calculating p-values to compare to a significance level of 0.05 in order to determine whether to reject the null hypothesis of no effect for each factor.

Uploaded by

Tony McDoan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
111 views28 pages

6.1 A. Estimate The Factor Effects. Which Effects Appear To Be Large?

The document reports the results of a statistical analysis with factors A, B, C, and D. Factors A, C, D, AB, AD, ABC, and ABD were found to have a significant effect on the response variable, while factors B, AC, BC, BD, CD, ACD, and BCD did not have a significant effect. The analysis included calculating p-values to compare to a significance level of 0.05 in order to determine whether to reject the null hypothesis of no effect for each factor.

Uploaded by

Tony McDoan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 28

6.1a.Estimatethefactoreffects.Whicheffectsappeartobelarge?

EffectB,C,andACappearstobelarge.

6.1b.
Letalpha=0.05.

FactorA
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.8837muchlargerthanalpha,thereforefailedtorejectnullhypothesis.Thereisno
significantdifferenceinfactorAeffect.

FactorB*
H0:
ThereisnosignificantdifferenceinfactorBeffect.
H1:
ThereissignificantdifferenceinfactorBeffect.

Pvalueis0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorBeffect.

FactorC*
H0:
ThereisnosignificantdifferenceinfactorCeffect.
H1:
ThereissignificantdifferenceinfactorCeffect.

Pvalueis0.0077smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorCeffect.
FactorAB
H0:
ThereisnosignificantdifferenceinfactorABfactoreffect.
H1:
ThereissignificantdifferenceinfactorABfactoreffect.

Pvalueis0.4681largerthanalpha,thereforefailedtorejectnullhypothesis.Thereisno
significantdifferenceinfactorABeffect.

FactorAC*
H0:
ThereisnosignificantdifferenceinfactorACfactoreffect.
H1:
ThereissignificantdifferenceinfactorACfactoreffect.

Pvalueis0.0012smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorACeffect.

FactorBC
H0:
ThereisnosignificantdifferenceinfactorBCfactoreffect.
H1:
ThereissignificantdifferenceinfactorBCfactoreffect.

Pvalueis0.2245islargerthanalpha,thereforefailedtorejectnullhypothesis.Thereisno
significantdifferenceinfactorBCeffect.

FactorABC
H0:
ThereisnosignificantdifferenceinfactorABCfactoreffect.
H1:
ThereissignificantdifferenceinfactorABCfactoreffect.

Pvalueis0.3483islargerthanalpha,thereforefailedtorejectnullhypothesis.Thereisno
significantdifferenceinfactorABCeffect.

6.1d

Thegraphlookslikeitformsastraightlinesoourassumptionthatthesampleisnormalisgood.

ShapiroWilkTest:
H0:thesamplescamefromaNormaldistribution
H1:thesamplesdidntcomefromaNormaldistribution

pval:0.0617>0.05,failedtorejectthenullhypothesis,thereforethesampledidcamefroma
Normaldistribution.

Thisgraphlooksverysymmetricandthereisnoclearpatternsotheconstantvariance
assumptionisnotviolated.

Thevarianceconstantassumptiondoesntseemtobeviolated,thelowandthehighlevelfor
factorAareroughlythesamelikewiseforfactorB.

ThevarianceconstantassumptiondoesntseemtobeviolatedinC,AB,AC,andBC.Thereis
nofunnelpattern.

ThevarianceconstantassumptiondoesntseemtobeviolatedinABCfactor.Thevarianceis
similartoeachother.Thereisnofunnelpattern.

6.e

Threefactors/interactionsaregraphedbecausetheyaretheoneswithsignificantdifferences.

Lookingatthesegraphs,wewouldliketohaveahighleveltoolgeometry(factorB)becauseit
increasesthelifeofamachinetool.Wewouldalsoliketohavealowcuttingspeed(factorA)
andahighlevelcuttingangle(factorC).

6.5a.

Letalpha=0.05.

FactorA*
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorAeffect.

FactorB*
H0:
ThereisnosignificantdifferenceinfactorBeffect.
H1:
ThereissignificantdifferenceinfactorBeffect.

Pvalueis0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorBeffect.

FactorAB*
H0:
ThereisnosignificantdifferenceinfactorABeffect.
H1:
ThereissignificantdifferenceinfactorABeffect.

Pvalueis0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorABeffect.

6.5b.

Thegraphlookslikeitformsastraightlinesoourassumptionthatthesampleisnormalisgood.

ShapiroWilkTest:
H0:thesamplescamefromaNormaldistribution
H1:thesamplesdidntcomefromaNormaldistribution

pval:0.8611>0.05,failedtorejectthenullhypothesis,thereforethesampledidcamefroma
Normaldistribution.

Thisgraphlooksverysymmetricandthereisnoclearpatternsotheconstantvariance
assumptionisnotviolated.

ThevarianceconstantassumptiondoesntseemtobeviolatedinA,B,andABgraph.Thereis
nofunnelpattern.

6.5c.

WewantalowlevelAwhichisasmallbitsizeof1/16inch.Thespeeddoesntmakemuchofa
differencewhenfactorAisatlowlevelbutwecanhaveBtobehighlevelwhichis90rpm.

6.7a

Theestimatedeffectaregiveninthesecondcolumnandthefirstcolumnlisttherespective
factors.

6.7b

FactorsA,C,D,AB,AD,ABC,ABDandABCDhassignificanteffectonyieldsincethePvalues
forthesefactorsarelessthan0.05.

Letalpha=0.05.

FactorA*
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorAeffect.

FactorB
H0:
ThereisnosignificantdifferenceinfactorBeffect.
H1:
ThereissignificantdifferenceinfactorBeffect.

Pvalueis0.1984largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBeffect.

FactorC*
H0:
ThereisnosignificantdifferenceinfactorCeffect.
H1:
ThereissignificantdifferenceinfactorCeffect.

Pvalueis0.0143smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorCeffect.

FactorD*
H0:
ThereisnosignificantdifferenceinfactorDeffect.
H1:
ThereissignificantdifferenceinfactorDeffect.

Pvalueis0.001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
Deffect.

FactorAB*
H0:
ThereisnosignificantdifferenceinfactorABeffect.
H1:
ThereissignificantdifferenceinfactorABeffect.

Pvalueis0.0007smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorABeffect.

FactorAC
H0:
ThereisnosignificantdifferenceinfactorACeffect.
H1:
ThereissignificantdifferenceinfactorACeffect.

Pvalueis0.4923largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorACeffect.

FactorAD*
H0:
ThereisnosignificantdifferenceinfactorADeffect.
H1:
ThereissignificantdifferenceinfactorADeffect.


Pvalueis0.0399smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorADeffect.
FactorBC
H0:
ThereisnosignificantdifferenceinfactorBCeffect.
H1:
ThereissignificantdifferenceinfactorBCeffect.

Pvalueis0.5733largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBCeffect.

FactorBD
H0:
ThereisnosignificantdifferenceinfactorBDeffect.
H1:
ThereissignificantdifferenceinfactorBDeffect.

Pvalueis0.8504largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBDeffect.

FactorCD
H0:
ThereisnosignificantdifferenceinfactorCDeffect.
H1:
ThereissignificantdifferenceinfactorCDeffect.

Pvalueis0.1038largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorCDeffect.

FactorABC*
H0:
ThereisnosignificantdifferenceinfactorABCeffect.
H1:
ThereissignificantdifferenceinfactorABCeffect.

Pvalueis<0.0001smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorABCeffect.

FactorABD*
H0:
ThereisnosignificantdifferenceinfactorABDeffect.
H1:
ThereissignificantdifferenceinfactorABDeffect.

Pvalueis0.0002smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorABDeffect.

FactorACD
H0:
ThereisnosignificantdifferenceinfactorACDeffect.
H1:
ThereissignificantdifferenceinfactorACDeffect.

Pvalueis0.3522largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorACDeffect.

FactorBCD
H0:
ThereisnosignificantdifferenceinfactorBCDeffect.
H1:
ThereissignificantdifferenceinfactorBCDeffect.


Pvalueis0.3522largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBCDeffect.

FactorABCD*
H0:
ThereisnosignificantdifferenceinfactorABCDeffect.
H1:
ThereissignificantdifferenceinfactorABCDeffect.

Pvalueis0.0241smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferencein
factorABCDeffect.

6.7c

6.7c
Plottheresidualsversusthepredictedyieldandonanormalprobabilityscale.
Doesthe
residualanalysisappearsatisfactory?

Thislooksgoodthereisnopatternhereanditisformingabandaroundthehorizontalline.

Yesthisdoesappeartobesatisfactory,it
lookslikeastraightlinearlinewithaslope,
soourassumptionthatthesampleis
normalisgood.

ShapiroWilkTest:
H0:thesamplescamefromaNormaldistribution
H1:thesamplesdidntcomefromaNormaldistribution

pval:0.2989>0.05,failedtorejectthenullhypothesis,thereforethesampledidcamefroma
Normaldistribution.

Ourconstantvarianceassumptionisnotviolatedthereisnofunnelorfanningpatterninanyof
thesegraphsbelow.

6.7e)Twothreefactorinteractions,
ABC
and
ABD
,apparentlyhavelargeeffects.Drawacubeplotinthe
factors
A,B
,and
C
withtheaverageyieldsshownateachcorner.Repeatusingthefactors
A,B
,and
D
.Do
thesetwoplotsaidindatainterpretation?Wherewouldyourecommendthattheprocessberunwith
respecttothefourvariables?

Seehandwrittenfordrawing.

Fromthecube,themeansofthefactorlevelsareshownonthecornersofthecubeplot.
Lowlevelsareleftfrontandrightfrontfortheleftcube.Thehighlevelisthebackright
andfrontrightintheleftcube.Therightcubethehighlevelonfrontandbackonthe
rightside.


7.2FactoreffectsandANOVAtable

Alltheeffects(A,B,AB)seemstobelarge.

Letalpha=0.05.
FactorA*
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.000muchsmallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorAeffect.

FactorB*
H0:
ThereisnosignificantdifferenceinfactorBeffect.
H1:
ThereissignificantdifferenceinfactorBeffect.

Pvalueis0.000smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorBeffect.

FactorAB*
H0:
ThereisnosignificantdifferenceinfactorABeffect.
H1:
ThereissignificantdifferenceinfactorABeffect.
Pvalueis0.000smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificant
differenceinfactorABeffect.


7.2residuals

Thegraphlookslikeitformsastraightlinesoour
assumptionthatthesampleisnormalisgood.

ShapiroWilkTest:
H0:thesamplescamefromaNormaldistribution
H1:thesamplesdidntcomefromaNormaldistribution

pval:0.5109>0.05,failedtorejectthenullhypothesis,thereforethesampledidcamefroma
Normaldistribution.

Thereisnopatternheresoourassumptionof
variancebeingconstantisnotviolated.

Ourconstantvarianceassumptionisnotviolatedthereisnofunnelorfanningpattern.

7.5Design


7.5Analyze

ThemaineffectsthatarelargeareA,D,AB,ABC,andABD.

Letalpha=0.05.

FactorA*
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.006smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
Aeffect.

FactorB
H0:
ThereisnosignificantdifferenceinfactorBeffect.
H1:
ThereissignificantdifferenceinfactorBeffect.

Pvalueis0.641largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBeffect.

FactorC
H0:
ThereisnosignificantdifferenceinfactorCeffect.
H1:
ThereissignificantdifferenceinfactorCeffect.

Pvalueis0.641largerthanalpha,thereforefailedtorejectnullhypothesis.Thereisnosignificantdifference
infactorCeffect.

FactorD*
H0:
ThereisnosignificantdifferenceinfactorDeffect.
H1:
ThereissignificantdifferenceinfactorDeffect.

Pvalueis0.041smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
Deffect.

FactorAB
H0:
ThereisnosignificantdifferenceinfactorABeffect.
H1:
ThereissignificantdifferenceinfactorABeffect.

Pvalueis0.053largerthanalpha,thereforefailedrejectnullhypothesis.Thereisnosignificantdifferencein
factorABeffect.

FactorAC
H0:
ThereisnosignificantdifferenceinfactorACeffect.
H1:
ThereissignificantdifferenceinfactorACeffect.

Pvalueis0.753largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorACeffect.

FactorAD
H0:
ThereisnosignificantdifferenceinfactorADeffect.
H1:
ThereissignificantdifferenceinfactorADeffect.

Pvalueis0.081largerthanalpha,thereforefailedtorejectnullhypothesis.Thereisnosignificantdifference
infactorADeffect.

FactorBC
H0:
ThereisnosignificantdifferenceinfactorBCeffect.
H1:
ThereissignificantdifferenceinfactorBCeffect.

Pvalueis0.452largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBCeffect.

FactorBD
H0:
ThereisnosignificantdifferenceinfactorBDeffect.
H1:
ThereissignificantdifferenceinfactorBDeffect.

Pvalueis0.377largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorBDeffect.


FactorCD
H0:
ThereisnosignificantdifferenceinfactorCDeffect.
H1:
ThereissignificantdifferenceinfactorCDeffect.

Pvalueis0.1038largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorCDeffect.

FactorABC*
H0:
ThereisnosignificantdifferenceinfactorABCeffect.
H1:
ThereissignificantdifferenceinfactorABCeffect.

Pvalueis0.026smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
ABCeffect.

FactorABD*
H0:
ThereisnosignificantdifferenceinfactorABDeffect.
H1:
ThereissignificantdifferenceinfactorABDeffect.

Pvalueis0.047smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
ABDeffect.

7.6design

7.6analysis

Letalpha=0.05.

FactorA*
H0:
ThereisnosignificantdifferenceinfactorAeffect.
H1:
ThereissignificantdifferenceinfactorAeffect.

Pvalueis0.000smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
Aeffect.

FactorD*
H0:
ThereisnosignificantdifferenceinfactorDeffect.
H1:
ThereissignificantdifferenceinfactorDeffect.

Pvalueis0.019smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
Deffect.

FactorAB*
H0:
ThereisnosignificantdifferenceinfactorABeffect.
H1:
ThereissignificantdifferenceinfactorABeffect.

Pvalueis0.019smallerthanalpha,thereforerejectnullhypothesis.Thereisasignificantdifferenceinfactor
ABeffect.

FactorABCD
H0:
ThereisnosignificantdifferenceinfactorABCDeffect.
H1:
ThereissignificantdifferenceinfactorABCDeffect.

Pvalueis0.093largerthanalpha,thereforefailedtorejectthenullhypothesis.Thereisnosignificant
differenceinfactorABCDeffect.

7.21

Factorsthatareconfounded:ABCD,ABEF,ACE,BDE,CDEF,BCF,ADF.

You might also like