/
Blocking and Confounding in Factorial Design of Experiments  Montgomery Chapter   Blocking Blocking and Confounding in Factorial Design of Experiments  Montgomery Chapter   Blocking

Blocking and Confounding in Factorial Design of Experiments Montgomery Chapter Blocking - PDF document

kittie-lecroy
kittie-lecroy . @kittie-lecroy
Follow
583 views
Uploaded On 2015-03-09

Blocking and Confounding in Factorial Design of Experiments Montgomery Chapter Blocking - PPT Presentation

within block Must do some sort of incomplete block analysis If you do not certain eects confounded Confounding two eects are indistinguishable May sacrice certain eects thought to be small design makes setup simple 241 Confounding in with only 2 blo ID: 43164

within block Must

Share:

Link:

Embed:

Download Presentation from below link

Download Pdf The PPT/PDF document "Blocking and Confounding in Factorial De..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

BlockingandConfoundinginFactorialDesignofExperiments-MontgomeryChapter7BlockinginFactorialDesignsForRCBD,eachcombinationrunineachblock4EUsperblock8EUsperblockRandomizerunorderwithinblockSupposeyoucannotrunallcomb.withinblockMustdosomesortofincompleteblockanalysisIfyoudonot,certaine®ectsconfoundedConfounding:twoe®ectsareindistinguishableMay\sacri¯ce"certaine®ectsthoughttobesmalldesignmakesset-upsimpleConfoundinginwithonly2blocksBlocksassumedtoallow2Firstconsider2factorial(2combsperblk)Possiblepairings1(1)andbtogetheraandabtogether2(1)andatogetherbandabtogether3(1)andabtogetheraandbtogether1E®ectofA,(ab+a-b-(1))/2,isblockdi®erence2E®ectofB,(ab-a+b-(1))/2,isblockdi®erenceBothhaveamaine®ectconfoundedwithblock3E®ectofAB,(ab-a-b+1)/2,isblockdi®erenceAllowsformaine®ectestimates(blkscancelout)ChoiceofConfoundingFactorsCommontoconfoundhighestorderinteractionCanuse+tabletodetermineblksFortwofactor,recallthefollowingtableABABSymbol --+(1)+--a-+-b+++ab Useconfoundingcolumntodetermineblks+'sinblk1and'sinblk2ConsiderthreefactorABCABACBCABCSymbol ---+++-(1)+----++a-+--+-+b++-+---ab--++--+c+-+-+--ac-++--+-bc+++++++abc Bestassignmentwouldbea,b,c,abctogetherCanestimateallbutthreefactorinteraction FactorialinFourBlocksFourblockseachcontaining2UsefulinsituationswhereMustselecttwoe®ectstoconfoundWillresultinathirdconfoundedfactorConsider6factorfactorialrunin4blocksof16EUsBlock1usesABC+andDEF+Block2usesABC+andDEF-Block3usesABC-andDEF+Block4usesABC-andDEF-Resultsin(ABC)(DEF)=ABCDEFconfoundedABCDandDEFABCEFconfoundedABandABEFEFconfoundedCanextendto8and16blksTable7-8summarizesthesedesigns(pg298)PartialConfoundingCanreplicateblockingdesignConfounddi®erente®ectseachreplicationAllowsestimationofalle®ectsConfoundede®ectsbasedonnonconfoundedreplicatesCanuseYates'Algorithmforallnonconfoundede®ectsSeeExample7-3(pg300)\*Example7-3*\datacool;inputblockfact1fact2fact3y;1-1-1-1-3111-12procglm;classfact1fact2fact3block;modely=blockfact1|fact2|fact3;ClassLevelsValuesFACT121-1FACT221-1FACT321-1BLOCK41234DependentVariable:YSourceDFSumofSquaresFValuePr�FModel1074.250000009.900.0103Error53.75000000CorrectedTotal1578.00000000SourceDFTypeISSFValuePr�FBLOCK33.500000001.560.3101FACT1136.0000000048.000.0010FACT2120.2500000027.000.0035FACT1*FACT210.500000000.670.4513FACT3112.2500000016.330.0099FACT1*FACT310.250000000.330.5887FACT2*FACT311.000000001.330.3004FACT1*FACT2*FACT310.500000000.670.4513Considera2factorialrunin4blocksEachreplicatewillresultin3confoundede®ectsConsider4replicatesfor32totalobservationsReplicate1:ConfoundBCandACReplicate2:ConfoundBCandABCReplicate3:ConfoundACandABCReplicate4:ConfoundABandABCThreereplicatestoestimateA,B,andCTworeplicatestoestimateAB,AC,andBCOnereplicatetoestimateABC Replicate1-AB,AC,BCConfoundedBlk1Blk2Blk3Blk475(1)89ab61a30b100abc73c45bc54acReplicate2-A,BC,ABCConfoundedBlk1Blk2Blk3Blk460(1)47a1b26ac34bc81abc35c52abReplicate3-B,AC,ABCConfoundedBlk1Blk2Blk3Blk458(1)48a18b68ab42ac52c82abc32bcReplicate4-C,AB,ABCConfoundedBlk1Blk2Blk3Blk447(1)34a50c37ac57ab4b80abc27bcSincenoe®ectisestimatedfromallreplications(excepterror),wewillcomputethee®ectsumsassociatedwitheachreplicateandcombinetheappropriateinformation.Thefollowingarethecolumn3sumsusingYates'Algorithmforeachofthereplicates.OnlythesumsthatarenotconfoundedwiththeparticularreparepresentedE®ectRep1Rep2Rep3Rep4 (1)5273364003368080AB-120120171616AC-BC--4040 (81+80+80) (1+0+0) (120+120) (16+16+17) (0+0) (40+40) 12 UsingSASoptionsnocenterps=50ls=80;datanew;inputreplblkabcresp;110007511111100121108912001734300150431118044101374401127procglm;classreplblkabc;modelresp=replblk(repl)a|b|c;SASOutputDependentVariable:RESPSumofMeanSourceDFSquaresSquareFValueModel2217128.718750778.57812528028.81Error90.2500000.027778CorrectedTotal3117128.968750SourceDFTypeISSMeanSquareFValueREPL33040.09375001013.364583336481.13BLK(REPL)127568.3750000630.697916722705.13A12420.04166672420.041666787121.50B10.04166670.04166671.50A*B13600.00000003600.000000099999.99C1100.0416667100.04166673601.50A*C10.00000000.00000000.00B*C1400.0000000400.000000014400.00A*B*C10.12500000.12500004.50SourceDFTypeIIISSMeanSquareFValueREPL33040.09375001013.364583336481.13BLK(REPL)12653.906250054.49218751961.72A12420.04166672420.041666787121.50B10.04166670.04166671.50A*B13600.00000003600.000000099999.99C1100.0416667100.04166673601.50A*C10.00000000.00000000.00B*C1400.0000000400.000000014400.00A*B*C10.12500000.12500004.50