Lecture16 - Lecture15 Stat3507.5 ConfidenceIntervalsfor...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Lecture15 Stat3507.5 ConfidenceIntervalsfor TwoPopula>onMeans LookingAhead... Today: ComparingTwoPopula>onMeans Topic1:LargeSamples Topic2:SmallSamples Topic3:PairedData WednesdayMarch2,2011 Lab3onConfidenceIntervals Duedate:FridayMarch11,2011 FridayMarch4,2011 Quiz[bringyourcalculators] 7.3:Comparingtwopopula>ons(largesamples) Examples: Domenhaveahighermeanmetabolismthan women? IsthemeanIQofPurduesimilartothemeanIQ ofIndianaU.? DoDuracellAAbaZerieslastlongerthan EnergizerAAbaZersonaverage? Nota>onfortwopopula>ons Population Population mean Population standard deviation 1 2 1 2 1 2 TwoSampleSe]ngs TwoSRSfromtwopopula>ons: SRSofsizen1isselectedfromthe1stpopula>on SRSofsizen2isselectedfromthe2ndpopula>on n1 andn2arebothlarge(30+),andcanbeofdifferentvalues Population Sample size Sample mean Sample standard deviation 1 2 n1 n2 s1 s2 Example4 Frompage307,wewanttocomparethelife>me (inhours)oftwobrandsofbaZeries: Whichbrandhasalongerlife>me?Byhow much? Whatsta>s>careyouusing? Whatistherelevantparameter? SAMPLINGDISTRIBUTION FORTHEDIFFERENCEOFMEANS Themean: Thestandarddevia>on: C.L.T.atworkagain Thesamplingdistribu>onofthedifferenceof meansisnormalifbothsamplesizesarelarge So, Howtostandardize? Recall, Similarly, Whenn1andn2arelarge,wecanreplacethe's withs's. TWOSAMPLECONFIDENCEINTERVAL FORTHEDIFFERENCEOFMEANS ThisCIisvalidforalldistribu>onsforthetwo popula>onswhenn1andn2arebothlarge Example4(cont'd) Frompage307,wewanttocomparethe life>me(inhours)oftwobrandsofbaZeries: Constructa95%confidenceintervalforthe differencebetweenthetwobrandsof baZeries. Again,makesureyoucaninterprettheinterval! Twosampleconfidenceintervalforthe differenceofmeans Zcrit(95%,doublesided)=? Twosampleconfidenceintervalforthe differenceofmeans Zcrit(95%,doublesided)=? 0.380.69 Ex32onpage312 !me beginning end Sample size 40 40 Sample mean 143.7 131.7 s 21.2 20.9 Find95%confidenceintervalforthetruemeandifference Ex32onpage312 !me beginning end Sample size 40 40 Sample mean 143.7 131.7 s 21.2 20.9 Find95%confidenceintervalforthetruemeandifference Interpretthe95%CI Yes,thereisevidencesugges>ngthatwhen thequizzesaretakenaffectsthetruemean scores.Infact,wewouldes>matethat studentstakingquizzesatthebeginningof classwouldscore,onaverage,between3and 21pointshigheronthefinalexam. Weare95%confidentinthisconclusion. 7.5Twosampletconfidenceintervals Atleastonesamplebeingsmall Twoindependentsamples FormulaforCI: df n 1 What is the degree of freedom? DegreeofFreedom df = (s (s 2 1 2 1 n1 + s n 2 ) 2 2 2 2 2 n1 ) n1 -1 (s + 2 2 n2 ) n 2 -1 Ifthedfisnotaninteger,rounddowntobe conserva>ve(I.E.df=9.86,use9) Example6 Metabolismratesof12randomwomenand 7randommenweremeasured. Finda95%CIforthedifferenceinmean metabolismbetweenmenandwomen Solu>on 364.0(tcrit)*89.77 Needtocalculatedegreeoffreedomtofindtcritval DegreeofFreedom df = (s (s 2 1 2 1 n1 + s n 2 ) 2 2 2 2 2 n1 ) n1 -1 (s + 2 2 n2 ) n 2 -1 Df=12.6rounddown,usedf=12 So, t cri!cal value (two sided 95% CI, df=12) = 2.179 Solu>on 364.02.179*89.77 =364.0195.6=(559.6,168.4) ACIfromPairedData Osen>mes,dataiscollectedinpairswhich createstheillusionoftwosamples,althoughin realitythereisreallyonlyonesample. Example--examsscoresforSTAT350 Obs Exam1 Exam2 1 74 87 2 89 86 3 83 79 ... Whyisthisconsideredonlyonesample? PairedData--treatlikeonesample Takethedifferenceofthescores,thenstudythe "differences"asasinglesample Obs Exam1 Exam2 Difference 1 74 82 8 2 89 86 3 3 83 79 4 ... Findthemeanofthedifferences. Findthestandarddevia>onofthedifferences. Example7 Supposeasampleofn studentsweregivenadiagnos>c testbeforeandagaincomple>ngamodule.Herewehave n1=n2=10students,thedataisasfollows: Finda95%confidenceintervalforthedifferences. Solu>on: Paired Data: n1=n2=10 Finda95%CIforthedifferences. Degreeoffreedom=9 Tcri>calvalue=2.262 Solu>on: Paired Data: n1=n2=10 Finda95%confidenceintervalforthedifferences. Degreeoffreedom=9Tcri>calvalue=2.262 Mean =1.600000 Std Devia!on = 2.95146 95%CI:1.6+/2.262*2.95/sqrt(10) =(1.62.1,1.6+2.1)=(0.5,3.7) ...
View Full Document

This note was uploaded on 02/06/2012 for the course STAT 350 taught by Professor Staff during the Spring '08 term at Purdue University-West Lafayette.

Ask a homework question - tutors are online