spss因子分析实例.docx

一.探讨目的：为了探讨农夫收入，我们选取了其中7种主要影峋因素，包括财政用于农业的支出的比重,其次、三产业从业人数占全社会从业人数的比(%)，非农村人口比玳，乡村从业人员占农村人口的比至(%),农业总产值占农林牧总产值的比重(%).农作物播种面积(千公顷)，农村用电量(亿千瓦时)°(数据见最终一页)二.探讨变量：在经济生活中，依据以上分析，我们在影响农夫收入因索中引入7个变量。即设置变量:xl-财政用于农业的支出的比重,x2-其次、三产业从业人数占全社会从业人数的比重，x3-非农村人口比重，x4-乡村从业人员占农村人口的比重，x5-农业总产值占农林牧总产值的比重,x6-农作物播种面积，x7农村用电最。一、探讨方法：SPSS中的因子分析。详细操作步骤(1)定义变量：XI-财政用于农业的支出的比重K2-其次、三产业从业人数占全社会从业人数的比重，x3-非农村人口比顶，X4-乡村从业人员占农村人口的比重，x5-农业总产值占农林牧总产值的比重，x6-农作物播种面积，x7农村用电量。(2)导入数据：(3)变此标准化AnaIyZC-DCSCriPtiVeStatistics-Dcscriptives一一-一W一一fliW'一0*Utitied2(DaUSetl-IBMSPSSSutisticsDataEditorFiIoEditVioWDataTransformAnabzeDicedMarketingGraphsUtJMiesAdd-onswin(:今H增国1.RegOits>*n1Ul这3DeSCriPbVGStabstics>71FfAatienciesSTables0DesiphvsAi×pore.SCrosstabs.K次业CompareMeans)General1.inearModel»Generaiued1.inearModels>I4edModels千;12220298-188213BBRatio-口PFPlots.3-.98719Correlate因Q-QPlots4-.57784Regression».5-.491781.ogimear56-.37008NeuralNetworks4729509Classi>58-25453DimensionReduction69-22257SqIe>510-.18692NonparametricTests1勾选Savestandardizedvaluesasvariables保存变量，再点击ok,就完成了对变量的标准化。（3）因子分析Analyze-DimensionReductionFaction整FactorAnalysisVanaWes夕年份"财政用于农业的支出/第二、三产业从业人."非农村人。比至（%）1.夕多村从曳人员占农村“农业电厂值占农樽牧夕农作物播加面收（千"农村用电艮（亿千瓦/Z88fe（W政用于夕Zsre第二、=.Zscore:非农村人.夕Zscore（乡村从业.夕ZSCore（农业总产夕Zscore农rm播8Zscore:农村用电SeiecbonVanaMeDesiptNes.ExtractionRotation.ScoresOptions.邠j幽Re寂MQan8lMHelpJ点击右侧的Description选项，选择Statistics选项组中的initialsolution,勾选CorrelationMatrix选项组中的Coefficients和KMOandBartleltstestofsphericity,点击Continueo制FactorAnalysis:OescriptivesrStatisticsUnivariatedesc11ptivesInitialsolutionrCofreatioMatrixCoefficientsinverseHSignificancelevelsReproducedJ5DeterminantAntnmageW0andBaitleirstestofsphericityContinueCancelHeIP点击右侧Extraction选项，其中MethOd选Principalcomponents,AnaIyZe选择Correlationmatrix,Display中选择Unrotatedfactorsolution,EXtraCt如图，点击COntinue.点击右侧Rotation选项，勾选Method选项组中的Varimax,Display中的两个选项都勾选，点击COntinue。&FactorAnalysis:RotationMethodONoneOQuaiUmax®VanmaxOEquamaxODirectONiminOPromaxDettaoKappa4DisplayRotatedsolution1.oadingpot(三)MaximumiterationstorConvergence:恒continueJCanCelHelp点击右侧SCOres,如图勾选，点击点击COntinUe。&FactorAnalysisFactorScoresXSaveasvariablesMethod®Regression。BaitlettOAnderson-RubinDisplayfactorscorecoefficientmatrixContinueCanCelHelP最终点击OPtions,默认*JFactorAnalysis:OptionsrMissingValuesExcludecasesIislwiseOExcludecasesBailWlSeOReplacewithmeanrCoeffientDisplayFormatSo<tedt>ysizeSuppresssmallcoef11etsAbsolutevaluebelow（4）结果分析1 KMOandBartletfs的检除结果图KMOandBartlett'sTestKaiser-Meyer-OlkinMeasureofSamplingAdequacy.725Bartlett'sTestofSphericityApprox.Chi-Square255.159df21Sig.000可以从今表中看出KMo统计以为0.725,大于最低标准，说明适合做因子分析，Bartlet环形检验，p<0.001,适合做因子分析。2 .主成分列表TotalVarianceExplainedComponentInitialEigenvaluesExtractionSumsofSquared1.oadingsRotationSumsofSquared1.oadingsTota1%ofVarianceCumulative%Total%ofVarianceCumulative%Total%ofVarianceCumulative%15.92O84.57284.5725.92084.57284.5723.30847.26147.2612.6539.33093.902.6539.33093.9023.26546.64193.9023.2493.55997.4624.1261.79899.2595.042.59599.8546.008.10899.9627.003.038100.000ExtractionMethod:PrincipalComponentAnalysis.可以从今表中看出前2个主成分特征值较大，它们的累积贡献率达到了93.902%,故选择前2个公共因子。3 .公因子方差比结果图CommunalitiesInitialExtractionZSCore(财政用于农业的支出的比重)1.000.906Zscore:其次、三产业从业人数占全社会从业人数的比重(%)1.000.940Zscore:非农村人口比重()1.000.979ZSCOre(乡村从业人员占农村人口的比重)1.000,977ZSCOre（农业总产值占农林牧总产值的比重）1.000.943Zscore:农作物播种面积（千公顷）1.000.909Zscore:农村用电量（亿千瓦时）1.000.918ExtractionMethod:PrincipalComponentAnalysis.结果显示，每一个指标变最的共性方羌都在09以上，说明这2个公共因子能够很好地反应原始各项指标变墙的绝大部分内容C4 .融荷做点图ComponentPlotInRotatedSpaceZX1.f1,”人口比ftl5HM.AAta<ntaZ&H用电TQlZ表作物七户工用I'公UOZ农aIt占我帆WfNwOZttfttIHOIKItK-1OSOQ0.5I(OComponent1从载荷散点图可以看出，第一公共因子能很好说明变景XI-财政用于农业的支出的比重，变量x5-农业总产值占农林牧总产值的比重，其次公共因子能很好地说明变及x2-其次、三产业从业人数占全社会从业人数的比重,x3-非农村人Il比重，x4-乡村从业人员占农村人口的比重，x6-农作物播种面积，x7农村用电量。5 .旋转后的因子荷图ComponentScoreCoefficientMatrixComponent12ZSCore（财政用于农业的支出的比重）.507-.697Zscore:其次、三产业从业人数占全社会从业人数的比重（）.120.112Zscore:非农村人口比重（）.170.066ZSCQre（乡村从业人员占农村人口的比重）.072.164ZSCOre（农业总产值占农林牧总产值的比重）.026-.257Zscore:农作物播种面积（千公顷）.691-.510Zscore:农村用电盘（亿千瓦时）.247-.022ExtractionMethod:PrincipalComponentAnalysis.RotationMethod:VarimaxwithKaiserNormalization.ComponentScores.经过旋转后，农作物播种面积（千公顷）、农村用电业（亿千瓦时）在因子一上方较大栽荷，财政用于农业的支出的比重、农业总产值占农林牧总产值的比重咋因子二上有较大载荷。故因子一可称为农业基本发展条件，因子二可称为政府支持状况。6历年农夫收入总得分降序表其中F=fl*84.57293.902+f2*9.33093.902年份iFlf2总分FII0.23231.3386214920041.46067141.08001.2253414220051.241375119981.44755-1.02581.2018006519990.88995-0.04300.797252111520000.833040.280990.77818891620010.798860.426520.76186470520020.567540.851630.59576687220030.296131.36620.40245098519970.355990.158990.33641629519960.1410.0230.12927564919860.0712-2.97824-0.2317890231991-0.3565-1-0.496-0.3703965931995-0.536810.53338-0.4304770921992-0.4608-0.2466-0.439580369031994-0.687930.39726-0.5801067091990-0.70907-0.2978-0.6682086251993-0.782350.24344-0.6804286281987-0.88133-1.7363-0.96628789261989-1.231950.22253-1.0874344581988-2.456461.00764-2.112270813数据:其次、三产年份财政用于农业的支出的比重业从业人数乡村从业人农业总产值占农作物播种农村用电非农村人口比占全社会从员占农村人农林牧总产值面积(千公量(亿千瓦重()业人数的比的比重的比重顷)时)重(%)198613.4329.517.9236.0179.99150104.07253.1198712.231.319.3938.6275.63146379.53320.819887.6637.623.7145.969.25143625.87508.919899.4239.926.2149.2362.75146553.93790.519909.9839.926.4149.9364.66148362.27844.5199110.2640.326.9450.9263.09149585.8963.2199210.0541.527.4651.5361.51149007.11106.919939.4943.627.9951.8660.07147740.71244.919949.245.728.5152.1258.22148240.61473.919958.4347.829.0452.4158.43149879.31655.719968.8249.530.4853.2360.57152380.61812.719978.350.131.9154.9358.23153969.21980.1199810.6950.233.3555.8458.03155705.72042.219998.2349.934.7857.1657.53156372.812173.4520007.755036.2259.3355.68156299.852421.320017.715037.6660.6255.24155707.862610.7820027.175039.0962.0254.51154635.512993.420037.1250.940.5363.7250.08152414.963432.9220049.6753.141.7665.6450.05153552.553933.0320057.2255.242.9967.5949.72155487.734375.7