統(tǒng)計學(xué)實驗分析-SPSS

重慶工商大學(xué)數(shù)學(xué)與統(tǒng)計學(xué)院《統(tǒng)計應(yīng)用軟件》課程實驗報告

實驗項目利用SPSS統(tǒng)計軟件，對數(shù)據(jù)進行分析。實驗日期周一，3-5節(jié)實驗地點80608實驗?zāi)康倪\用SPSS軟件對居民儲蓄調(diào)查數(shù)據(jù)進行基本統(tǒng)計分析，T檢驗，單因素分析，多因素方差分析，相關(guān)分析，偏相關(guān)分析，以及曲線估計。二項logistic回歸分析。實驗內(nèi)容：第一步用統(tǒng)計軟件SPSS打開居民儲蓄調(diào)查數(shù)據(jù)。第二步進行分析。一．根據(jù)提供數(shù)據(jù)，對居民調(diào)查數(shù)據(jù)進行基本統(tǒng)計分析。1．選擇菜單Analyze-Descriptivestatistics-Frequencies,出現(xiàn)下圖所示2．得出結(jié)論FrequencyTable職業(yè)FrequencyPercentValidPercentCumulativePercentValid國家機關(guān)248.58.58.5商業(yè)服務(wù)業(yè)5419.119.127.7文教衛(wèi)生186.46.434.0公交建筑業(yè)155.35.339.4經(jīng)營性公司186.46.445.7學(xué)校155.35.351.1一般農(nóng)戶3512.412.463.5種糧棉專業(yè)戶41.41.464.9種果菜專業(yè)戶103.53.568.4工商運專業(yè)戶3412.112.180.5退役人員176.06.086.5金融機構(gòu)3512.412.498.9現(xiàn)役軍人31.11.1100.0Total282100.0100.0年齡FrequencyPercentValidPercentCumulativePercentValid20歲以下41.41.41.420~35歲14651.851.853.235~50歲9132.332.385.550歲以上4114.514.5100.0Total282100.0100.0BarChart得出結(jié)論：1按職業(yè)進行分類。例如，職業(yè)為國家機關(guān)的人數(shù)為24人，占總?cè)藬?shù)的8.5%；一般農(nóng)戶的人數(shù)為35人，占總?cè)藬?shù)的2.4%。按年齡劃分，20歲以下的人共有4人，占總?cè)藬?shù)的1.4%；20~35歲共有146人，占總?cè)藬?shù)的51.8%。因為職業(yè)和年齡在變量中，無去ishishuju，因為百分比與有效百分比總是相同的。2根據(jù)柱形圖，本次被調(diào)查者的年齡分布情況是：20-35歲最多，其次是35-50歲，50歲以上，最少的人數(shù)是20歲以下。當(dāng)選擇PIECHART時，得出結(jié)論如下圖。PieChart2．選擇菜單Analyze-Descriptivestatistics-Frequencies,出現(xiàn)下圖所示得出結(jié)論：Descriptives上圖表明，二．對數(shù)據(jù)進行T檢驗分析。1．選擇Analyze-comparemeans-one-samplesTtest如下圖。選擇Option2得出結(jié)論：T-TestOne-SampleStatisticsNMeanStd.DeviationStd.ErrorMean存(取)款金額2824738.0910945.569651.799One-SampleTestTestValue=2000tdfSig.(2-tailed)MeanDifference95%ConfidenceIntervaloftheDifferenceLowerUpper存(取)款金額4.201281.0002738.0891455.064021.12該問題應(yīng)采用雙尾檢驗，因此比較a/2和p/2。由于p小于a，因此拒絕零假設(shè)。認(rèn)為出乎總體一次平均村會計呢與2000元有顯著差異。95%的置信區(qū)間告之，有95%的把握認(rèn)為儲戶總體平均一次在3455.06–6021.12元之間。2000元不包含在置信區(qū)間內(nèi)，也證實了上述推斷。三．對數(shù)據(jù)進行多因素方差檢驗。1選擇菜單Analyze-generallinearmodel-univariate于是出現(xiàn)窗口均值比較?？刂谱兞拷换プ饔脠D形分析2．得出結(jié)論UnivariateAnalysisofVarianceBetween-SubjectsFactorsValueLabelN收入水平1300元以下502300~800元1643800~1500元5041500元以上18物價水平1過高452偏高1793正常58TestsofBetween-SubjectsEffectsDependentVariable:什么合算SourceTypeIIISumofSquaresdfMeanSquareFSig.CorrectedModel3.787a10.3791.621.100Intercept160.1291160.129685.487.000a41.1563.3851.650.178a8.0862.043.184.832a4*a82.1555.4311.845.104Error63.305271.234Total612.000282CorrectedTotal67.092281a.RSquared=.056(AdjustedRSquared=.022)上圖表明，第一列是對觀測變量總變差分解的說明；第二列是觀測變量變差分解的結(jié)果。第三列表明自由度；第四列是均方；第五列是F檢驗統(tǒng)計量的觀測值；第六列是檢驗統(tǒng)計量的概率P值。可以看出，觀測變量的總變差為67.092。它由4部分組成，分別是收入水平a4=1.156,物價水平a8=.086,由收入水平和物價水平交互作用引起的變差a4*a8=2.155,由隨機因素引起的變差error=63.305。這些變差除以各自的自由度后，得到各自的均方，并可計算出各F檢驗統(tǒng)計量的觀測值和在一定自由度下的概率P值（即第六列）。若顯著性水平α=0.05,由于Fa4,Fa8的概率p值小于顯著性水平α，則應(yīng)拒絕零假設(shè)?？梢哉J(rèn)為收入水平和物價水平總體均值存在顯著差異，對什么合算的效應(yīng)不同時為0，各自不同的水平給什么合算帶來了顯著影響。同時由于Fa4,Fa8的概率p值小于顯著性水平α，則應(yīng)拒絕零假設(shè)，可以認(rèn)為不同的收入水平和物價水平對什么合算帶來了交互作用，不同的收入水平與物價水平會影響到購買什么合算。CustomHypothesisTestsContrastResults(KMatrix)收入水平DeviationContrastaDependentVariable什么合算Level1vs.MeanContrastEstimate.008HypothesizedValue0Difference(Estimate-Hypothesized).008Std.Error.084Sig..92195%ConfidenceIntervalforDifferenceLowerBound-.156UpperBound.173Level2vs.MeanContrastEstimate-.020HypothesizedValue0Difference(Estimate-Hypothesized)-.020Std.Error.065Sig..75895%ConfidenceIntervalforDifferenceLowerBound-.149UpperBound.108Level3vs.MeanContrastEstimate-.036HypothesizedValue0Difference(Estimate-Hypothesized)-.036Std.Error.080Sig..65195%ConfidenceIntervalforDifferenceLowerBound-.193UpperBound.121a.Omittedcategory=4上圖分別表明收入水平前三個水平下買什么合算的均值檢驗結(jié)果，省略了第四水平的檢驗結(jié)果，檢驗值是總體均值?？梢钥闯觯谝环N收入水平下的什么合算的均值與檢驗值差0.008，標(biāo)準(zhǔn)誤差為0.084，檢驗統(tǒng)計量的概率p值為0.0921，差值的95%置信區(qū)間的下限和上限分別是-0.156和0.173。分析結(jié)論是，第一種收入水平下的合算均值與檢驗值不存在顯著差異。同理，第二種和第三種也不存在顯著差異。所以可得出，在三種情況下，都不存在顯著差異。TestResultsDependentVariable:什么合算SourceSumofSquaresdfMeanSquareFSig.Contrast.0633.021.090.966Error63.305271.234ProfilePlots上圖表明從收入水平過高，偏高，正常的漸變期間，對購買力有很大程度上的影響。即可以同上述存在明顯交互作用達(dá)成一致結(jié)論。四．對數(shù)據(jù)進行偏相關(guān)數(shù)據(jù)分析。1.選擇菜單Analyze-correlate-partial2.選擇參與分析的變量到variable3.選擇一個或多個控制變量到controllingfor框中選擇OptionZero-ordercorrelation表示輸出零階偏相關(guān)系數(shù)。2．得出結(jié)論PartialCorrCorrelationsControlVariables存(取)款金額收入水平物價趨勢-none-a存(取)款金額Correlation1.000.184.014Significance(2-tailed)..002.821df0280280收入水平Correlation.1841.000-.001Significance(2-tailed).002..983df2800280物價趨勢Correlation.014-.0011.000Significance(2-tailed).821.983.df2802800物價趨勢存(取)款金額Correlation1.000.184Significance(2-tailed)..002df0279收入水平Correlation.1841.000Significance(2-tailed).002.df2790a.Cellscontainzero-order(Pearson)correlations.五．對數(shù)據(jù)計算相關(guān)系數(shù)1選擇菜單analyze-correlate-bivariate2選擇options—cross—productdeviationsandcovariances表示輸出各變量的離差平方和，樣本方差，兩變量的叉積離差和協(xié)方差3得出結(jié)論CorrelationsCorrelations收入情況存(取)款金額物價水平收入情況PearsonCorrelation1-.090.046Sig.(2-tailed).131.439SumofSquaresandCross-products101.872-167118.5324.723Covariance.363-594.728.017N282282282存(取)款金額PearsonCorrelation-.0901-.008Sig.(2-tailed).131.898SumofSquaresandCross-products-167118.5323.367E10-14216.152Covariance-594.7281.198E8-50.591N282282282物價水平PearsonCorrelation.046-.0081Sig.(2-tailed).439.898SumofSquaresandCross-products4.723-14216.152102.401Covariance.017-50.591.364N282282282上圖表明，收入情況與存取款金額的簡單相關(guān)系數(shù)為-0.090，與物價水平的簡單相關(guān)系數(shù)為0.439。它們的相關(guān)系數(shù)檢驗的概率p分別為0.131和0.439。當(dāng)顯著性水平α為0.05或0.01時，都接受相關(guān)系數(shù)檢驗的零假設(shè)，認(rèn)為兩總體不存在線性關(guān)系。六．對數(shù)據(jù)進行曲線估計選擇菜單Analyze-regression-curveestimation。2得出結(jié)論CurveFitModelDescriptionModelNameMOD_1DependentVariable1收入情況Equation1Quadratic2CubicIndependentVariable存(取)款金額ConstantIncludedVariableWhoseValuesLabelObservationsinPlots存（?。┛罘N類ToleranceforEnteringTermsinEquationsCaseProcessingSummaryNTotalCases282ExcludedCasesa0ForecastedCases0NewlyCreatedCases0a.Caseswithamissingvalueinanyvariableareexcludedfromtheanalysis.VariableProcessingSummaryVariablesDependentIndependent收入情況存(取)款金額NumberofPositiveValues282282NumberofZeros00NumberofNegativeValues00NumberofMissingValuesUser-Missing00System-Missing00收入情況ModelSummaryRRSquareAdjustedRSquareStd.ErroroftheEstimate.146.021.014.598Theindependentvariableis存(取)款金額.ANOVASumofSquaresdfMeanSquareFSig.Regression2.16621.0833.031.050Residual99.706279.357Total101.872281Theindependentvariableis存(取)款金額.CoefficientsUnstandardizedCoefficientsStandardizedCoefficientstSig.BStd.ErrorBeta存(取)款金額-1.855E-5.000-.337-2.396.017存(取)款金額**22.048E-10.000.272..(Constant)2.080.04348.500.000CubicModelSummaryRRSquareAdjustedRSquareStd.ErroroftheEstimate.165.027.017.597Theindependentvariableis存(取)款金額.ANOVASumofSquaresdfMeanSquareFSig.Regression2.7843.9282.603.052Residual99.089278.356Total101.872281Theindependentvariableis存(取)款金額.CoefficientsUnstandardizedCoefficientsStandardizedCoefficientstSig.BStd.ErrorBeta存(取)款金額-3.277E-5.000-.596-2.467.014存(取)款金額**28.193E-10.0001.090..存(取)款金額**3-5.130E-15.000-.598..(Constant)2.104.04745.121.000七．二項Logistic回歸分析1選擇菜單Analyze-Regression-Binarylogistic。添加變量。2單擊categorical按鈕指定如何生成虛擬變量，選擇變量3選擇Options和save4得出結(jié)論LogisticRegressionCaseProcessingSummaryUnweightedCasesaNPercentSelectedCasesIncludedinAnalysis28199.6MissingCases1.4Total282100.0UnselectedCases0.0Total282100.0a.Ifweightisineffect,seeclassificationtableforthetotalnumberofcases.DependentVariableEncodingOriginalValueInternalValue城鎮(zhèn)戶口0農(nóng)村戶口1CategoricalVariablesCodingsFrequencyParametercoding(1)(2)(3)(4)(5)存（?。┛罘N類三年以上定期501.000.000.000.000.000三年以下定期59.0001.000.000.000.000活期135.000.0001.000.000.000定活兩便3.000.000.0001.000.000活期工資帳戶22.000.000.000.0001.000各類信用卡12.000.000.000.000.000現(xiàn)金不存的原因日常生活用錢156.000.000.000生意周轉(zhuǎn)金341.000.000.000歲制購買中意商品57.0001.000.000購買國庫券34.000.0001.000Block0:BeginningBlockClassificationTablea,bObservedPredicted戶口PercentageCorrect城鎮(zhèn)戶口農(nóng)村戶口Step0戶口城鎮(zhèn)戶口1990100.0農(nóng)村戶口820.0OverallPercentage70.8a.Constantisincludedinthemodel.b.Thecutvalueis.500VariablesintheEquationBS.E.WalddfSig.Exp(B)Step0Constant-.887.13145.6461.000.412VariablesnotintheEquationScoredfSig.Step0Variablesa415.6481.000a620.4645.001a6(1)3.4451.063a6(2)12.0701.001a6(3)10.5991.001a6(4)1.2501.264a6(5).4811.488a1232.3893.000a12(1)23.6201.000a12(2).2841.594a12(3)12.8881.000OverallStatistics68.3149.000Block1:Method=EnterOmnibusTestsofModelCoefficientsChi-squaredfSig.Step1Step79.1119.000Block79.1119.000Model79.1119.000ModelSummaryStep-2LoglikelihoodCox&SnellRSquareNagelkerkeRSquare1260.207a.245.350a.Estimationterminatedatiterationnumber20becausemaximumiterationshasbeenreached.Finalsolutioncannotbefound.HosmerandLemeshowTestStepChi-squaredfSig.14.0007.780ContingencyTableforHosmerandLemeshowTest戶口=城鎮(zhèn)戶口戶口=農(nóng)村戶口TotalObservedExpectedObservedExpectedStep112625.6590.3412622827.63511.3652932624.34623.6542844441.71668.2845052022.21696.7842961515.68498.3162471517.3651310.6352881615.3511414.64930999.0282827.97237ClassificationTableaObservedPredicted戶口PercentageCorrect城鎮(zhèn)戶口農(nóng)村戶口Step1戶口城鎮(zhèn)戶口1891095.0農(nóng)村戶口513137.8OverallPercentage78.3a.Thecutvalueis.500VariablesintheEquationBS.E.WalddfSig.Exp(B)Step1aa4-1.085.25418.1761.000.338a618.3475.003a6(1)1.286.9671.7711.1833.619a6(2)1.843.9573.7081.0546.312a6(3).304.932.1061.7441.355a6(4)-18.11120993.578.0001.999.000a6(5).4721.076.1931.6611.604a1228.3803.000a12(1)2.330.51420.5111.00010.274a12(2)-.083.384.0471.829.920a12(3)-2.2841.0564.6751.031.102Constant.2491.035.0581.8101.283a.Variable(s)enteredonstep1:a4,a6,a12.Stepnumber:1ObservedGroupsandPredictedProbabilities80++||||F||R60++E||Q|2|U|1|E40+1+N|1|C|1|Y|1|20+122+|111222||1121111111212||111111111121211211112221222|Predicted---------+---------+---------+---------+---------+---------+---------+---------+---------+----------Prob:0.1.2.3.4.5.6.7.8.91Group:1111111111111111111111111111111111111111111111111122222222222222222222222222222222222222222222222222PredictedProbabilityisofMembershipfor農(nóng)村戶口TheCutValueis.50Symbols:1-城鎮(zhèn)戶口2-農(nóng)村戶口EachSymbolRepresents5Cases.八．單因素分析1選擇analyse-comparemeans-oneway-anova得出結(jié)論OnewayANOVASumofSquaresdfMeanSquareFSig.收入水平BetweenGroups(Combined)1.7983.5991.006.390LinearTermUnweighted1.20311.2032.019.156Weighted.0011.001.002.965Deviation1.7972.8991.508.223WithinGroups165.606278.596Total167.404281存(取)款金額BetweenGroups(Combined)5.657E831.886E81.584.193LinearTermUnweighted3.753E813.753E83.152.077Weighted1.282E811.282E81.077.300Deviation4.375E822.188E81.837.161WithinGroups3.310E102781.191E8Total3.367E10281ContrastCoefficientsContrast年齡20歲以下20~35歲35~50歲50歲以上11-.50-.5ContrastTestsContrastValueofContrastStd.ErrortdfSig.(2-tailed)收入水平Assumeequalvariances1-.61.392-1.553278.122Doesnotassumeequalvariances1-.61.297-2.0473.370.123存(取)款金額Assumeequalvariances19426.495540.3701.701278.090Doesnotassumeequalvariances19426.4912365.123.7623.009.501PostHocTestsMultipleComparisonsDependentVariable(I)年齡(J)年齡MeanDifference(I-J)Std.ErrorSig.95%ConfidenceIntervalLowerBoundUpperBound收入水平Scheffe20歲以下20~35歲-.644.391.440-1.74.4635~50歲-.654.394.433-1.76.4650歲以上-.573.404.571-1.71.5620~35歲20歲以下.644.391.440-.461.7435~50歲-.010.1031.000-.30.2850歲以上.071.136.966-.31.4535~50歲20歲以下.654.394.433-.461.7620~35歲.010.1031.000-.28.3050歲以上.081.145.958-.33.4950歲以上20歲以下.573.404.571-.561.7120~35歲-.071.136.966-.45.3135~50歲-.081.145.958-.49.33LSD20歲以下20~35歲-.644.391.101-1.41.1335~50歲-.654.394.098-1.43.1250歲以上-.573.404.157-1.37.2220~35歲20歲以下.644.391.101-.131.4135~50歲-.010.103.923-.21.1950歲以上.071.136.605-.20.3435~50歲20歲以下.654.394.098-.121.4320~35歲.010.103.923-.19.2150歲以上.081.145.579-.21.3750歲以上20歲以下.573.404.157-.221.3720~35歲-.071.136.605-.34.2035~50歲-.081.145.579-.37.21存(取)款金額Scheffe20歲以下20~35歲8327.8155530.039.520-7226.4723882.1035~50歲7305.9515574.425.633-8373.1822985.0850歲以上10525.1595715.751.337-5551.4826601.8020~35歲20歲以下-8327.8155530.039.520-23882.107226.4735~50歲-1021.8651457.357.921-5120.963077.2350歲以上2197.3431928.597.730-3227.207621.8935~50歲20歲以下-7305.9515574.425.633-22985.088373.1820~35歲1021.8651457.357.921-3077.235120.9650歲以上3219.20820