第六章方差分析ppt课件.ppt

《第六章方差分析ppt课件.ppt》由会员分享，可在线阅读，更多相关《第六章方差分析ppt课件.ppt（47页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、选择选择分析分析菜单下菜单下均值比较均值比较下的下的因素方差分析因素方差分析选项，选项，则打开则打开单因素方差分析单因素方差分析对话框。对话框。1 选择选择变量变量 数学数学2 单击单击按钮按钮5单击单击ok按钮按钮3 选择选择变量变量 组别组别4 单击单击按钮按钮组间ANOVA3320.77821660.38915.595.0001597.00015106.4674917.77817Between GroupsWithin GroupsTotalSum ofSquaresdfMean SquareFSig.平方和自由度均方F值P值组内总ANOVA数学Sum of SquaresdfMean

2、SquareFSig.Between Groups2620.33321310.1678.271.004Within Groups2376.16715158.411Total4996.50017单击单击options 则打开则打开options对话框对话框输出描述性统计量输出描述性统计量要求进行方差齐性检验要求进行方差齐性检验要求绘制均值分布图要求绘制均值分布图DescriptivesDescriptives数学693.174.9562.02387.9798.378899680.3315.4886.32364.0896.595699664.8312.6405.16051.5778.1055881

3、879.4416.3363.85071.3287.575599012TotalNMeanStd. DeviationStd. ErrorLower BoundUpper Bound95% Confidence Interval forMeanMinimumMaximumTest of Homogeneity of VariancesTest of Homogeneity of Variances数学2.183215.147LeveneStatisticdf1df2Sig.由P值可知方差齐性。012组 别60708090Mean of 数学单击单击Post Hoc按钮，按钮，则打开则打开Post

4、 Hoc对话框对话框方差不齐时方差不齐时方差齐性时方差齐性时DescriptivesDescriptives数学693.174.9562.02387.9798.378899680.3315.4886.32364.0896.595699664.8312.6405.16051.5778.1055881879.4416.3363.85071.3287.575599012TotalNMeanStd. DeviationStd. ErrorLower BoundUpper Bound95% Confidence Interval forMeanMinimumMaximumTest of Homogene

5、ity of VariancesTest of Homogeneity of Variances数学2.183215.147LeveneStatisticdf1df2Sig.Multiple ComparisonsMultiple ComparisonsDependent Variable: 数学LSD12.8336.865.081-1.8027.4728.333*6.865.00113.7042.97-12.8336.865.081-27.471.8015.500*6.865.039.8730.13-28.333*6.865.001-42.97-13.70-15.500*6.865.039-

6、30.13-.87(J) 组别120201(I) 组别012MeanDifference(I-J)Std. ErrorSig.Lower BoundUpper Bound95% Confidence IntervalThe mean difference is significant at the .05 level.*. *代表存在差异代表存在差异 a男男990b女女880c男男990d男男890e女女940f男男900g男男791h女女561i男男891j男男991k女女701l男男891m女女552n男男502o女女672p男男672q女女562r男男562选择分析分析菜单下的一般线性模

7、型一般线性模型下单因变量多因素方差分析单因变量多因素方差分析选项打开单因变量多因素方差分析对话框打开单因变量多因素方差分析对话框因变量因变量自变量自变量单击事后检验方差齐性检验描述性分析单击B Be et tw we ee en n- -S Su ub bj je ec ct ts s F Fa ac ct to or rs s男11女766612性别012组别Value LabelN各个控制变量水平下观察个案的个数D De es sc cr ri ip pt ti iv ve e S St ta at ti is st ti ic cs sDependent Variable: 数学94.2

8、55.500489.008.165457.678.622382.3617.3511191.004.243263.009.899259.336.658369.4315.936793.174.956680.3315.488658.506.950677.3317.57318组别012Total012Total012Total性别男女TotalMeanStd. DeviationNL Le ev ve en ne e s s T Te es st t o of f E Eq qu ua al li it ty y o of f E Er rr ro or r V Va ar ri ia an nc c

9、e es sa aDependent Variable: 数学.339512.879Fdf1df2Sig.Tests the null hypothesis that the error variance ofthe dependent variable is equal across groups.Design: Intercept+性别+组别+性别 * 组别a. T Te es st ts s o of f B Be et tw we ee en n- -S Su ub bj je ec ct ts s E Ef ff fe ec ct ts sDependent Variable: 数学

10、4605.917a5921.18317.163.00095235.260195235.2601774.340.000351.1571351.1576.542.0253295.57721647.78830.700.000599.8432299.9225.588.019644.0831253.674112898.000185250.00017SourceCorrected ModelIntercept性别组别性别 * 组别ErrorTotalCorrected TotalType III Sumof SquaresdfMean SquareFSig.R Squared = .877 (Adjust

11、ed R Squared = .826)a. 看性别、组别，性别看性别、组别，性别*组别这三行组别这三行性别存在主效应性别存在主效应，组别存在主效应组别存在主效应。性别与组别存在交互作用性别与组别存在交互作用。性别主效应组别主效应性别与组别交互作用Multiple ComparisonsMultiple ComparisonsDependent Variable: 数学LSD12.83*4.230.0103.6222.0534.67*4.230.00025.4543.88-12.83*4.230.010-22.05-3.6221.83*4.230.00012.6231.05-34.67*4.2

12、30.000-43.88-25.45-21.83*4.230.000-31.05-12.62(J) 组别120201(I) 组别012MeanDifference(I-J)Std. ErrorSig.Lower BoundUpper Bound95% Confidence IntervalBased on observed means.The mean difference is significant at the .05 level.*. 0组与组与1组、组、2组的成绩存在显著差异，组的成绩存在显著差异，1组与组与2组也存在组也存在显著差异。显著差异。EstimatesEstimatesD

13、ependent Variable: 数学94.2503.66386.269102.23189.0003.66381.01996.98157.6674.23048.45166.88391.0005.18079.713102.28763.0005.18051.71374.28759.3334.23050.11768.549组别012012性别男女MeanStd. ErrorLower BoundUpper Bound95% Confidence IntervalUnivariate TestsUnivariate TestsDependent Variable: 数学14.083114.083.

14、262.618644.0831253.674901.3331901.33316.793.001644.0831253.6744.16714.167.078.785644.0831253.674ContrastErrorContrastErrorContrastError组别012Sum ofSquaresdfMean SquareFSig.Each F tests the simple effects of 性别 within each level combination of the othereffects shown. These tests are based on the linea

15、rly independent pairwisecomparisons among the estimated marginal means.性别的简单效应检验性别的简单效应检验P Pa ai ir rw wi is se e C Co om mp pa ar ri is so on ns sDependent Variable: 数学3.2506.345.618-10.57417.074-3.2506.345.618-17.07410.57426.000*6.345.00112.17639.824-26.000*6.345.001-39.824-12.176-1.6675.982.785-1

16、4.70011.3671.6675.982.785-11.36714.700(J) 性别女男女男女男(I) 性别男女男女男女组别012MeanDifference(I-J)Std. ErrorSig.aLower BoundUpper Bound95% Confidence Interval forDifferenceaBased on estimated marginal meansThe mean difference is significant at the .050 level.*. Adjustment for multiple comparisons: Least Signifi

17、cant Difference (equivalent to noadjustments).a. EstimatesEstimatesDependent Variable: 数学94.2503.66386.269102.23191.0005.18079.713102.28789.0003.66381.01996.98163.0005.18051.71374.28757.6674.23048.45166.88359.3334.23050.11768.549性别男女男女男女组别012MeanStd. ErrorLower BoundUpper Bound95% Confidence Interva

18、lUnivariate TestsUnivariate TestsDependent Variable: 数学2571.12921285.56423.952.000644.0831253.6741319.0482659.52412.288.001644.0831253.674ContrastErrorContrastError性别男女Sum ofSquaresdfMean SquareFSig.Each F tests the simple effects of 组别 within each level combination of the othereffects shown. These

19、tests are based on the linearly independent pairwisecomparisons among the estimated marginal means.组别的简单效应检验组别的简单效应检验Pairwise ComparisonsPairwise ComparisonsDependent Variable: 数学5.2505.180.331-6.03716.53736.583*5.595.00024.39248.775-5.2505.180.331-16.5376.03731.333*5.595.00019.14243.525-36.583*5.59

20、5.000-48.775-24.392-31.333*5.595.000-43.525-19.14228.000*7.326.00212.03843.96231.667*6.688.00017.09546.238-28.000*7.326.002-43.962-12.0383.6676.688.594-10.90518.238-31.667*6.688.000-46.238-17.095-3.6676.688.594-18.23810.905(J) 组别120201120201(I) 组别012012性别男女MeanDifference(I-J)Std. ErrorSig.aLower BoundUpper Bound95% Confidence Interval forDifferenceaBased on estimated marginal meansThe mean difference is significant at the .050 level.*. Adjustment for multiple comparisons: Least Significant Difference (equivalent to noadjustments).a.