1、統(tǒng)計學(xué)例題操作分析（僅供參考） 1.單樣本t檢驗（樣總t檢驗）P55L4-7 SPSS操作主要結(jié)果（1）進行正態(tài)性檢驗Tests of NormalityKolmogorov-SmirnovaShapiro-WilkStatisticdfSig.StatisticdfSig.M物質(zhì).1449.200*.9649.840a. Lilliefors Significance Correction*. This is a lower bound of the true significance.（2） 進行單樣本t檢驗One-Sample TestTest Value = 63.5 tdfSig.

2、(2-tailed)Mean Difference95% Confidence Interval of the DifferenceLowerUpperM物質(zhì)-45.1428.000-21.13333-22.2129-20.05382.配對t檢驗（異體配對/自身前后、左右配對）P57L4-8 SPSS操作主要結(jié)果（1） 先計算配對差值d再對d進行正態(tài)性檢驗Tests of NormalityKolmogorov-SmirnovaShapiro-WilkStatisticdfSig.StatisticdfSig.d.19310.200*.96310.819a. Lilliefors Signif

3、icance Correction*. This is a lower bound of the true significance.（2） 進行配對t檢驗Paired Samples CorrelationsNCorrelationSig.Pair 1對照 & 實驗10.075.838Paired Samples TestPaired DifferencestdfSig. (2-tailed)MeanStd. DeviationStd. Error Mean95% Confidence Interval of the DifferenceLowerUpperPair 1對照 - 實驗2.16

4、0001.32094.417721.215053.104955.1719.0013. 兩獨立樣本t檢驗（成組t檢驗） P59L4-9 SPSS操作主要結(jié)果（1）進行正態(tài)性和方差齊性檢驗Tests of Normality觀察對象Kolmogorov-SmirnovaShapiro-WilkStatisticdfSig.StatisticdfSig.轉(zhuǎn)鐵蛋白含量正常人.11612.200*.98612.998患者.14715.200*.92715.245a. Lilliefors Significance Correction*. This is a lower bound of the tru

5、e significance.Test of Homogeneity of VarianceLevene Statisticdf1df2Sig.轉(zhuǎn)鐵蛋白含量Based on Mean1.877125.183Based on Median1.127125.299Based on Median and with adjusted df1.127122.304.300Based on trimmed mean1.798125.192（2）再做成組t檢驗Independent Samples TestLevenes Test for Equality of Variancest-test for Eq

6、uality of Means95% Confidence Interval of the DifferenceFSig.tdfSig. (2-tailed)Mean DifferenceStd. Error DifferenceLowerUpper轉(zhuǎn)鐵蛋白含量Equal variances assumed1.877.1837.40225.00036.66004.952826.459546.8605Equal variances not assumed7.67524.795.00036.66004.776426.818646.50144.完全隨機設(shè)計資料ANOVA與多重比較 P69L5-2 S

7、PSS操作主要結(jié)果（1） 正態(tài)性和方差齊性檢驗Tests of Normality藥物Kolmogorov-SmirnovaShapiro-WilkStatisticdfSig.StatisticdfSig.E-SFC對照組.1799.200*.9479.653淫羊霍.13610.200*.95610.739黨參.19710.200*.92310.380黃芪.15110.200*.95210.692a. Lilliefors Significance Correction*. This is a lower bound of the true significance.Test of Homo

8、geneity of VariancesE-SFCLevene Statisticdf1df2Sig.2.601335.067（2） 單因素方差分析ANOVAE-SFCSum of SquaresdfMean SquareFSig.Between Groups1978.9443659.64877.789.000Within Groups296.800358.480Total2275.74438Multiple ComparisonsDependent Variable:E-SFC(I) 藥物(J) 藥物Mean Difference (I-J)Std. ErrorSig.95% Confide

9、nce IntervalLower BoundUpper BoundLSD對照組淫羊霍-20.067*1.338.000-22.78-17.35黨參-7.667*1.338.000-10.38-4.95黃芪-8.067*1.338.000-10.78-5.35淫羊霍對照組20.067*1.338.00017.3522.78黨參12.400*1.302.0009.7615.04黃芪12.000*1.302.0009.3614.64黨參對照組7.667*1.338.0004.9510.38淫羊霍-12.400*1.302.000-15.04-9.76黃芪-.4001.302.761-3.042.2

10、4黃芪對照組8.067*1.338.0005.3510.78淫羊霍-12.000*1.302.000-14.64-9.36黨參.4001.302.761-2.243.04Dunnett t (2-sided)a淫羊霍對照組20.067*1.338.00016.7923.34黨參對照組7.667*1.338.0004.3910.94黃芪對照組8.067*1.338.0004.7911.34*. The mean difference is significant at the 0.05 level.a. Dunnett t-tests treat one group as a control,

11、and compare all other groups against it.5.隨機區(qū)組（配伍組）設(shè)計資料兩因素ANOVA與多重比較 P72L5-3 SPSS操作主要結(jié)果Tests of Between-Subjects EffectsDependent Variable:治愈天數(shù)SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model43.467a67.24416.718.000Intercept1520.06711520.0673507.846.000年齡24.93346.23314.385.001療法18.53329

12、.26721.385.001Error3.4678.433Total1567.00015Corrected Total46.93314a. R Squared = .926 (Adjusted R Squared = .871)Pairwise ComparisonsDependent Variable:治愈天數(shù)(I) 療法(J) 療法Mean Difference (I-J)Std. ErrorSig.a95% Confidence Interval for DifferenceaLower BoundUpper Bound中西結(jié)合中醫(yī)-.600.416.188-1.560.360西醫(yī)-2.

13、600*.416.000-3.560-1.640中醫(yī)中西結(jié)合.600.416.188-.3601.560西醫(yī)-2.000*.416.001-2.960-1.040西醫(yī)中西結(jié)合2.600*.416.0001.6403.560中醫(yī)2.000*.416.0011.0402.960Based on estimated marginal meansa. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).*. The mean difference is sign

14、ificant at the .05 level.6.拉丁方設(shè)計資料三因素ANOVA P76L5-4 SPSS操作主要結(jié)果Tests of Between-Subjects EffectsDependent Variable:斑塊面積SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model1726.562a9191.84014.145.002Intercept.0631.06310773.461.000中藥1608.6883536.22939.538.000窩別18.68736.229.459.721飼料99.187333.0

15、622.438.163Error81.375613.563Total.00016Corrected Total1807.93715a. R Squared = .955 (Adjusted R Squared = .887)Pairwise ComparisonsDependent Variable:斑塊面積(I) 中藥(J) 中藥Mean Difference (I-J)Std. ErrorSig.a95% Confidence Interval for DifferenceaLower BoundUpper BoundAB-12.000*2.604.004-18.372-5.628C10.

16、000*2.604.0093.62816.372D-15.250*2.604.001-21.622-8.878BA12.000*2.604.0045.62818.372C22.000*2.604.00015.62828.372D-3.2502.604.259-9.6223.122CA-10.000*2.604.009-16.372-3.628B-22.000*2.604.000-28.372-15.628D-25.250*2.604.000-31.622-18.878DA15.250*2.604.0018.87821.622B3.2502.604.259-3.1229.622C25.250*2

17、.604.00018.87831.622Based on estimated marginal means*. The mean difference is significant at the .05 level.a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).7.2x2析因設(shè)計資料的ANOVA P77L5-5 SPSS操作主要結(jié)果Tests of Between-Subjects EffectsDependent Variable:膽固醇降

18、低量(mg%)SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model5298.917a31766.30641.077.000Intercept21252.083121252.083494.234.000a藥4144.08314144.08396.374.000b藥884.0831884.08320.560.002a藥 * b藥270.7501270.7506.297.036Error344.000843.000Total26895.00012Corrected Total5642.91711a. R Squared = .9

19、39 (Adjusted R Squared = .916)8.2x2交叉設(shè)計資料ANOVA P78L5-6 SPSS操作主要結(jié)果Tests of Between-Subjects EffectsDependent Variable:巴氏小胃鹽酸分泌量(mol/h)SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model1128.000a1386.7694.132.015Intercept8664.00018664.000412.571.000階段24.000124.0001.143.310狗號1008.0001191.636

20、4.364.014藥物96.000196.0004.571.058Error210.0001021.000Total10002.00024Corrected Total1338.00023a. R Squared = .843 (Adjusted R Squared = .639)9.組內(nèi)分組設(shè)計資料兩因素ANOVA P79L5-7 SPSS操作主要結(jié)果Tests of Between-Subjects EffectsDependent Variable:M物質(zhì)SourceType I Sum of SquaresdfMean SquareFSig.Corrected Model6.114a4

21、1.52912.778.000Intercept138.5391138.5391158.124.000產(chǎn)地5.85022.92524.452.000部位.2642.1321.104.363Error1.43512.120Total146.08917Corrected Total7.55016a. R Squared = .810 (Adjusted R Squared = .746)Pairwise ComparisonsDependent Variable:M物質(zhì)(I) 產(chǎn)地(J) 產(chǎn)地Mean Difference (I-J)Std. ErrorSig.a95% Confidence In

22、terval for DifferenceaLower BoundUpper Bound1.327*.210.000.8691.785.306.223.196-.181.792-1.327*.210.000-1.785-.869-1.021*.233.001-1.528-.514-.306.223.196-.792.1811.021*.233.001.5141.528Based on estimated marginal means*. The mean difference is significant at the .05 level.a. Adjustment for multiple

23、comparisons: Least Significant Difference (equivalent to no adjustments).Pairwise ComparisonsDependent Variable:M物質(zhì)(I) 部位(J) 部位Mean Difference (I-J)Std. ErrorSig.a95% Confidence Interval for DifferenceaLower BoundUpper Bound當(dāng)歸頭當(dāng)歸身-.266.180.164-.658.125當(dāng)歸尾-.192.309.546-.865.481當(dāng)歸身當(dāng)歸頭.266.180.164-.125

24、.658當(dāng)歸尾.075.298.807-.575.724當(dāng)歸尾當(dāng)歸頭.192.309.546-.481.865當(dāng)歸身-.075.298.807-.724.575Based on estimated marginal meansa. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).10.完全隨機設(shè)計兩組協(xié)方差分析 P103L6-10 SPSS操作主要結(jié)果（1） 用散點圖初步判斷兩比較組是否有直線趨勢（2） 用交互項來檢驗兩比較組總體直線回歸系數(shù)是否相等

25、，結(jié)果如下Tests of Between-Subjects EffectsDependent Variable:用藥后SourceType I Sum of SquaresdfMean SquareFSig.Corrected Model1389.299a3463.10021.357.000Intercept.2861.28611829.679.000組別326.9041326.90415.076.001用藥前1029.54111029.54147.479.000組別 * 用藥前32.854132.8541.515.230Error520.4152421.684Total.00028Corr

26、ected Total1909.71427a. R Squared = .727 (Adjusted R Squared = .693)（3）比較修正均數(shù)是否相等Tests of Between-Subjects EffectsDependent Variable:用藥后SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model1356.445a2678.22330.646.000Intercept14.025114.025.634.433組別251.3201251.32011.356.002用藥前1029.54111029.5

27、4146.521.000Error553.2692522.131Total.00028Corrected Total1909.71427a. R Squared = .710 (Adjusted R Squared = .687)11.隨機區(qū)組（配伍組）設(shè)計資料的協(xié)方差分析 P104L6-11 SPSS操作主要結(jié)果 （1）用散點圖初步判斷三組的進食量與體重增量是否都有直線趨勢（2）用交互項來檢驗三組總體直線回歸系數(shù)相等，結(jié)果如下Tests of Between-Subjects EffectsDependent Variable:體重增量SourceType I Sum of Squaresd

28、fMean SquareFSig.Corrected Model73210.103a164575.63137.601.000Intercept.5881.5881330.885.000飼料47809.096223904.548196.441.000窩別19137.032111739.73014.297.000進食量6197.42416197.42450.929.000飼料 * 進食量66.550233.275.273.764Error2312.07019121.688Total.76036Corrected Total75522.17235a. R Squared = .969 (Adjust

29、ed R Squared = .944)（3）比較修正均數(shù)是否相等Tests of Between-Subjects EffectsDependent Variable:體重增量SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model73143.552a145224.53946.126.000Intercept1708.43611708.43615.083.001飼料478.9122239.4562.114.146窩別3796.27511345.1163.047.014進食量6197.42416197.42454.715.00

30、0Error2378.62021113.268Total.76036Corrected Total75522.17235a. R Squared = .969 (Adjusted R Squared = .948)12.四格表的Pearson 2檢驗 P121L8-1 SPSS操作主要結(jié)果Chi-Square TestsValuedfAsymp. Sig. (2-sided)Exact Sig. (2-sided)Exact Sig. (1-sided)Pearson Chi-Square56.772a1.000.000.000Continuity Correctionb53.6191.000

31、Likelihood Ratio49.4681.000Fishers Exact TestLinear-by-Linear Association56.6211.000N of Valid Cases376a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 8.24.b. Computed only for a 2x2 table注： Pearson Chi-Square 適用條件 n40 且T5 Continuity Correctionb n40 且1T5 Fishers Exact

32、 Test n40 且T113.RxC表2檢驗與多重比較 P126 L8-4 SPSS操作主要結(jié)果Chi-Square TestsValuedfAsymp. Sig. (2-sided)Pearson Chi-Square9.596a6.143Likelihood Ratio9.7416.136Linear-by-Linear Association4.0151.045N of Valid Cases199a. 2 cells (16.7%) have expected count less than 5. The minimum expected count is 3.59.14.方表資料檢

33、驗（Pearson，McNemar，Kappa） P130L8-9 SPSS操作主要結(jié)果Chi-Square TestsValueExact Sig. (2-sided)McNemar Test.454aN of Valid Cases168a. Binomial distribution used.Symmetric MeasuresValueAsymp. Std. ErroraApprox. TbApprox. Sig.Measure of AgreementKappa.809.04510.500.000N of Valid Cases168a. Not assuming the null

34、 hypothesis.b. Using the asymptotic standard error assuming the null hypothesis.15.完全隨機設(shè)計兩組計量資料Wilcoxon秩和檢驗 P14216.完全隨機設(shè)計兩組等級資料Mann-Whitney U檢驗 P143L9-3 SPSS操作主要結(jié)果Test Statisticsb抑制率Mann-Whitney U15.000Wilcoxon W43.000Z-.857Asymp. Sig. (2-tailed).391Exact Sig. 2*(1-tailed Sig.).445aa. Not corrected

35、for ties.b. Grouping Variable: 葛根劑量17.完全隨機設(shè)計多組計量資料Kruskal-Wallis H 檢驗 P144L9-5 SPSS操作主要結(jié)果Test Statisticsb死亡率N15Median24.0000Chi-Square10.179adf2Asymp. Sig.006a. 6 cells (100.0%) have expected frequencies less than 5. The minimum expected cell frequency is 2.3.b. Grouping Variable: 藥物Test Statisticsa

36、,b死亡率Chi-Square9.740df2Asymp. Sig.008a. Kruskal Wallis Testb. Grouping Variable: 藥物18. 兩個大樣本等級資料Ridit分析 P15219. 多個大樣本等級資料Ridit分析L9-11 SPSS操作主要結(jié)果（1） 計算標(biāo)準(zhǔn)組（2） 計算標(biāo)準(zhǔn)組療效各等級R值（3） 計算對比組總體R均值的置信區(qū)間DescriptivesStatisticStd. ErrorPROPORTION of 療效 using RANKITMean.95% Confidence Interval for MeanLower Bound.Upp

37、er Bound.5% Trimmed Mean.Median.Variance.051Std. Deviation.Minimum.3611Maximum.9810Range.6199Interquartile Range.4532Skewness1.119.171Kurtosis-.611.34120.圓形分布資料兩角均數(shù)的比較 P163L10-5 (1)參數(shù)法Watson-William檢驗(適用于有集中趨勢的資料r 0.7)/非參數(shù)法Watsons U2檢驗；（2）正態(tài)性檢驗；（3）兩樣本平均角比較的t檢驗。21. 多應(yīng)變量多組Wilks lambda檢驗（多元方差分析） P231L1

38、6-3 SPSS操作結(jié)果Multivariate TestscEffectValueFHypothesis dfError dfSig.InterceptPillais Trace.987994.174a2.00026.000.000Wilks Lambda.013994.174a2.00026.000.000Hotellings Trace76.475994.174a2.00026.000.000Roys Largest Root76.475994.174a2.00026.000.000貧血類別Pillais Trace.5535.1594.00054.000.001Wilks Lambda

39、.5115.192a4.00052.000.001Hotellings Trace.8345.2094.00050.000.001Roys Largest Root.6388.614b2.00027.000.001a. Exact statisticb. The statistic is an upper bound on F that yields a lower bound on the significance level.c. Design: Intercept + 貧血類別Pairwise ComparisonsDependent Variable(I) 貧血類別(J) 貧血類別Me

40、an Difference (I-J)Std. ErrorSig.a95% Confidence Interval for DifferenceaLower BoundUpper Bound血紅蛋白AB-11.033*3.154.002-17.505-4.561C-.3333.362.922-7.2336.566BA11.033*3.154.0024.56117.505C10.700*3.494.0053.53017.870CA.3333.362.922-6.5667.233B-10.700*3.494.005-17.870-3.530紅細胞計數(shù)AB-.353.180.060-.723.016

41、C-.483*.192.018-.877-.089BA.353.180.060-.016.723C-.130.200.520-.540.280CA.483*.192.018.089.877B.130.200.520-.280.540Based on estimated marginal means*. The mean difference is significant at the .05 level.a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustmen

42、ts).22.重復(fù)測量資料的ANOVA P233L16-4 SPSS操作主要結(jié)果Multivariate TestsbEffectValueFHypothesis dfError dfSig.factor1Pillais Trace.85420.469a2.0007.000.001Wilks Lambda.14620.469a2.0007.000.001Hotellings Trace5.84820.469a2.0007.000.001Roys Largest Root5.84820.469a2.0007.000.001factor1 * 組別Pillais Trace.4492.857a2.

43、0007.000.124Wilks Lambda.5512.857a2.0007.000.124Hotellings Trace.8162.857a2.0007.000.124Roys Largest Root.8162.857a2.0007.000.124a. Exact statisticb. Design: Intercept + 組別 Within Subjects Design: factor1Mauchlys Test of SphericitybMeasure:MEASURE_1Within Subjects EffectMauchlys WApprox. Chi-Squared

44、fSig.EpsilonaGreenhouse-GeisserHuynh-FeldtLower-boundfactor1.5713.9212.141.700.909.500Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables is proportional to an identity matrix.a. May be used to adjust the degrees of freedom for the averag

45、ed tests of significance. Corrected tests are displayed in the Tests of Within-Subjects Effects table.b. Design: Intercept + 組別 Within Subjects Design: factor1Tests of Within-Subjects EffectsMeasure:MEASURE_1SourceType III Sum of SquaresdfMean SquareFSig.factor1Sphericity Assumed190.067295.03334.246

46、.000Greenhouse-Geisser190.0671.400135.78934.246.000Huynh-Feldt190.0671.818104.56634.246.000Lower-bound190.0671.000190.06734.246.000factor1 * 組別Sphericity Assumed24.867212.4334.480.029Greenhouse-Geisser24.8671.40017.7664.480.048Huynh-Feldt24.8671.81813.6804.480.033Lower-bound24.8671.00024.8674.480.067Error(factor1)Sphericity Assumed44.400162.775Greenhouse-Geisser44.40011.1983.965Huynh-Feldt44.40014.5413.053Lower-bound44.4008.0005.55023.多重線性回歸 P241L17-1 SPSS操作主要結(jié)果Model SummarycModelRR SquareAdjusted R SquareStd. Error