class13 Regression Analysis of Correlated Data-new-f2008

class13 Regression Analysis of Correlated Data-new-f2008

ID:37577616

大小:223.30 KB

页数:27页

时间:2019-05-25

class13 Regression Analysis of Correlated Data-new-f2008_第1页
class13 Regression Analysis of Correlated Data-new-f2008_第2页
class13 Regression Analysis of Correlated Data-new-f2008_第3页
class13 Regression Analysis of Correlated Data-new-f2008_第4页
class13 Regression Analysis of Correlated Data-new-f2008_第5页
资源描述:

《class13 Regression Analysis of Correlated Data-new-f2008》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库

1、TheRegressionAnalysisofCorrelatedDataTherearemanysituationsthatresultinsamplesthatarenotmadeupofindependentobservations•Familydata:Wecanexpectthatsubjectsinthesamefamilywillhaveobservationsthataremorealikewhencomparedtoeachotherthanwhencomparedtosubjectsinotherfamilies.•Datac

2、ollectedusingnon-randomsampling:Whenusingasamplingschemethatselectsseveralhospitalsfromwhichsubjectswillberecruited,subjectsatthesamehospitalwillshareunmeasuredcharacteristicsthatwouldnothappenifsubjectswereselectedrandomlyfromthegeneralpopulation.•Repeatedmeasures:Whentaking

3、repeatedmeasurementsfromsubjects,observationsthatarefromthesamesubjectwilltendtobemorealikethanwhencomparedtoothersubjectsobservations.Repeatedobservationscloserintimefromthesamesubjectmayalsobemorealikethanthosefurtherapartintime.2ConsequencesofNon-IndependentDataConsiderthi

4、sexperiment:•Youattendaconventionforidenticaltripletsthathas1,000setsofidenticaltriplets.•Youdecideyouhavetimetotake1,000systolicbloodpressuremeasurements.Youcaneither:1.Collectbloodpressuresfromallmembersof333-1/3setsoftripletsor,2.Takeonemeasurefromeachofthe1,000sets.Ifyouw

5、antaproperestimateofSBPattheconvention,areyoubetteroffdoing1or2?•Ifmembersofatriplettendtohavesimilarbloodpressures,yougetlittleadditionalinformationbymeasuringmorethanonememberofasibship.•Thetotalinformationcontentofyourdatawillbereducedifyoudon’thaveindependentobservations.

6、•Thestandarderrorsonyourestimateswillbetoosmallifyoudon’taccountforthecorrelations–yourdataisnotasgoodasyouthink!3Situation#2:SleepHoursinMenandWomen•Youaretestingthehypothesisthatmothersandfathersofnewbornswillgetsignificantlydifferenthoursofsleep,whereyoubelievethatmenwills

7、leepmorethanwomen.•Youcandothestudytwoways:1.Youcollect50couplesandcalculatethedifferenceinsleephoursbetweenthemotherandfatherofanewborn.Isthisdifferencesignificantlydifferentfromzero?[pairedt-testwith50pairs].2.Youaccumulate50mothersand50fathersofnewborns,withnoattempttorecr

8、uitcouples.Youthencalculatethemeannumberofhoursofsleepinmothersandin

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。