第9章_相关分析

《第9章_相关分析》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、第9章相关分析三Kfl一SPSSDataEditorHEllAnalyzeClillxle?IlndowgclpReportsDfifcriptiveStatisticsXabletCompareleansGeneralLinearI^dellizedModelsCorrelatePartial..・Distances・・・RegressionL^Clinear图9-1Correlate的级联菜单Bivariate和Partial等9.1双变量相关分析9.1.1数据图9-2北京市2003年SARS疫惜数据9.1.2利用相关

2、矩阵进行分析C^rrelatieasmt*AftWlr*A»FetrienCorrelation1000«Z3•336S&C000008M61et61wKiTObiRsrFtarsenCorrelation.9231000-304sicm“).000onVStSI61f&Correlation-33d-304I000Sic2-tailtd).006on61G161图9-3非典与最高气温的相关情况（截至双解除日）CerrelatBons新增3*5甘唧量低气豪毅増非典人35PttrsonCorrtltlion1.000923•

3、610Si<(2-U.O•.000000616161斬版個非Q病刮PttrsonCorrtlilion・*231000-566Si{G-tkiltd).000000V6161ei*iKnaPtaxsonCarreletion■・610-5M1000SiC12-tMUd).000000•2图9-4（截至双解除日）北京SARS疫情与每日故低气温的关系9.1.3缺失值的剔除法对相关矩阵的影响1.采用默认的双举剔除缺失值法及双尾概率法的输出结果Cerrelati•・■狷視1U丰•病洌当丟肅F«ers^aC^xr«l«ti9&S&c

4、.e-Uiled)N174-&13^(no73-380*00173・299・・.01172Pe«rs©aCcrrelationSic(2~tailed)N-61>»00073iT3T6却00072-.258**.03071員鬲气溫FtirscaCorrelationSicUZN-30O»t00173T6XwoT2173.42ROOO71当日逼差FtarsonCorrelationSi€(2-taiUd)N299»t：Oil72-2S8«030T1422"OOO71172*•-Correlationisatthe0.01le

5、velwCorroliseaatat006Itrel(2・ZiL@4)图9・5双举剔除缺失值法及双尾概率法的输出结杲2.采用单举剔除缺失值法和单尾概率法的输出结果Cerr«lalA・》1FearsenCorrelatioa1•60卯OOO•364M001007Corrtlaii9ASi<(1-ttiled)■603M.0001760M000・25“017PearsonCorrelationSic(l"teiled)-.33001760m143W000当日盪雯Fearson.CorrelationSic(1-tailed).

6、00?-25301T43Sm0001CorreL'donis軌aificglittheQ.01ierelQ-tcled).・Cerrelallonsi^nifleantetthe005levelI.LittwiitB=70图9・6单举剔除缺失值法利单尾概率法的输出结果9.1.4Correlation的对话框及其解释1.打开BivariateCorrelation对话框的步骤BivariateCorrelations◎统计时lol■Variables:0K。省市的編号[Id2]•匚、；l城历例阿。新增非典人败凹♦豪低nsai

7、ti]Paste。前增非翼人数累计2m气通问•医务人员累计非典*~T1■每日溫塑拠Reset。累计出険人敛[“]I-U—4新增死亡人数[V5]Cancel①累计新增死亡人数—e累计疑似非典病例卜」HelpCorrelationCoefficientsFPearson厂KendalKstau-b厂Spearman-TestofSignificanceOptions...eIwo-tailedrOne-tailedPBagsignificantcorrelations图9-7BivariateCorrelations对话框2

8、.对主对话框的说明3.统计量的选择Continue

9、BivariateCorrelations:Option»StatisticsrMeansandstandarddeviationsrCrossproductdeviationscovarimcceMissingVilue€(•Excludecases