浙大生物统计样卷2010-2011-A答案

Biostatistics余思扬31001002272012年6月-1-浙江大学2010–2011学年秋冬学期《生物统计(学)与实验设计》课程期末考试试卷—A卷Problem1(25points):OneexperimentisconductedforstudyingtheinfluenceofplantconditiononthecontentofNicotineinleafoftobacco.Twotobaccovarietiesandtwodifferentplantcultivations(fertilization,nofertilization)werearranged,2piecesofleavesweresampledforeachvariety.Theobtainedexperimentdataareshowninthefollowingtable.叶片1叶片2叶片1叶片2施肥施肥不施肥不施肥品种1951920471416品种21517151712111211(1)Whatisthisexperimentdesign?ThisisatypicalCrossedNestedDesign.(2)WriteouttheANOVAmodelofthisexperimentforanalysis;defineeachfactorinthemodel.Error:Leaf:Variety:ionFertilizat:Mean:2,12,12,12,1αnkjiYijknjikijjkjiijkn(3)FortheaboveANOVAmodel,writeouttheformulaofdegreefreedomandcorrespondingsumofsquareforfactorsinthemodel.SourceDegreeofFreedom(df)SumofSquares(SS)Aa-1=1aiiYYbcr12Bb-1=1bjjYYacr12C(B)b(c-1)=2bjckjkYYar112AB(a-1)(b-1)=1aibjjiijYYYYcr112AC(B)(a-1)b(c-1)=2aibjckjkijijkYYYYr1112Errorabc(r-1)=8aibjckrnijkijknYY11112Note:A,B,CandErrorarereferredtofactorα,β,γandErrorε.Anda,b,candrarethenumbertestedinthisdesignoffactorα,β,γandErrorε.浙江大学2010–2011学年秋冬学期《生物统计(学)与实验设计》课程期末考试试卷A卷-2-(4)WriteouttheSASprogramforanalysisofthisdata.(5)AccordingtothefollowingoutputofSASanalysis,drawappropriatestatisticalconclusion.FromtheSASresult,wecouldfindthatthePrFofmodelis0.18610.05,themodelisnotreachsignificantlevel,needtobeadjusted.AndfromTypeIIISSwecanfindthatonlythefactorthatsignificantisfertilization.Sotheconclusionisinthemodel,fertilizationissignificantbutthemodelitselfisnotreachthesignificantlevel,weneedtochangemodel.Andalsowecanfindinthismodel,alltheinteractionfactorsaremissing,sothechangeofmodelwecanaddalltheinteractionsasIwriteinQuestion(4)above,andwemaygetabetterresult.DATANicotine;INPUTFertilization$Variety$Leaf$Content@@;DATALINES;A1B1C19A1B1C25A2B1C119A2B1C220A1B1C14A1B1C27A2B1C114A2B1C216A1B2C115A1B2C217A2B2C115A2B2C217A1B2C112A1B2C211A2B2C112A2B2C211;PROCGLM;CLASSFertilizationVarietyLeaf;MODELContent=FertilizationVarietyLeaf(Variety)Fertilization*VarietyFertilization*Leaf(Variety);RANDOMLeaf(Variety)/TEST;CONTRAST'LeafinB1'Leaf(Variety)1-100;CONTRAST'LeafinB2'Leaf(Variety)001-1;MEANSFertilization/TUKEY;MEANSVariety/TUKEY;LSMEANSLeaf(Variety)/STDERRPDIFF;LSMEANSFertilization*Variety/STDERRPDIFF;LSMEANSFertilization*Leaf(Variety)/STDERRPDIFF;RUN;浙江大学2010–2011学年秋冬学期《生物统计(学)与实验设计》课程期末考试试卷A卷-3-Problem2(30points):Oneexperimentofricevarietyisconductedforstudyingtherelationshipofyieldandplantingdensity.Theexperimenthas2varieties(A1,A2)and3plantingdensities10(B1),20(B2),30(B3).Theobservationdataareshowninthefollowingtable.B1B2B3B1B2B3B1B2B3A1888776656A2998796876(1)Isitafactorialdesignornesteddesign?ItisatypicalTwo-wayFactorialDesign.(2)WriteouttheANOVAmodelforthisexperiment.Error:Density:Variety:Mean:3,2,13,2,12,1αkjiYijkijjiijk(3)Cantheinteractionofvarietyandplantingdensitybeanalyzed？Why?Yes.Becauseinthisexperimentwehadthreereplicates,sowecanconducttheanalysisofinteraction.(4)Assumethevarietyisfixed,plantingdensityisrandom;writeouttheSASprogramtoanalysisthisdata.(5)HowtousetheSAStotestthedifferencebetweenB3andtheaverageofB1andB2?WeshouldadaptLinearContrastcommandinthistest:AddthistermintoPROC:(6)Forthissetofexperimentdata,howtoconstructaregressionmodelofriceyieldonplantingdensity,sothatitcanbeusedtopredictriceyieldforotherplantingdensity.CONTRAST'Linear'Density11-2;DATARice;INPUTVariety$Density$Yield@@;DATALINES;A1B18A1B28A1B38A1B17A1B27A1B36A1B16A1B25A1B36A2B19A2B29A2B38A2B17A2B29A2B36A2B18A2B27A2B36;PROCGLM;CLASSVarietyDensity;MODELYield=Variety|Density;MEANSVariety/TUKEY;MEANSDensity/TUKEY;LSMEANSVariety*Density/STDERRPDIFF;RUN;浙江大学2010–2011学年秋冬学期《生物统计(学)与实验设计》课程期末考试试卷A卷-4-Problem3(15points):Thefollowingdataisabouttheheight(X1),weight(X2)andbodysurfacearea(Y)ofteninfants.MaleFemaleX1X2YX1X2Y5432446.25432117.350.52.251928.4532.252200.2512.52094.551.52.51906.256.53.52506.75131850.352321215131632.5(1)Howtotestthedifferenceinbodysurfaceareabetweenmaleandfemale?Inthisproblem,weshouldemploytheAnalysisofCovariance(ANOCOVA).(2)WriteouttheSASprogramfortheabovestatisticaltesting.(3)Forthefollowingresultsofanalysis,drawappropriatestatisticalconclusion.FromtheSASresult,wecouldfindthatthePrFofmodelis0.00140.01,themodelisverysignificant,andwecandofurtheranalysis.AndfromTypeIIISSwecanfindthatthefactorthatsignificantareHeightandSex.DATAInfants;INPUTX1X2SexY@@;DATALINES;543S12446.2543S22117.350.52.25S11928.4532.25S22200.2512.5S12094.551.52.5S21906.256.53.5S12506.7513S21850.3523S12121513S21632.5;PROCGLM;CLASSSex;MODELY=SexX1X2;LSMEANSSex/STDERRPDIFFADJUST=TUKEYETYPE=3;RUN;浙江大学2010–2011学年秋冬学期《生物统计(学)与实验设计》课程期末考试试卷A卷-5-Sotheconclusionisinthemodel,HeightandSexaresignificant.Andalsowecanfindinthismodel,thefactorWeightisnotreachsignificantlevel.Andfromallabovethefinalconclusionisthebodysurfaceisrelatedwithheightandsex.Problem4(30points):ThesimulationdataofQTL