[统计学秩和检验.ppt

Biostatistics * 例题 (假设不满足参数检验的要求) Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 例题 Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 多组等级比较的检验假设 建立检验假设 H0 ：各组总体的等级分布相同； H1 ：各组总体的等级分布不同或不全相同。 ? =0.05。 计算检验统计量H值 Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 多组等级比较的检验假设 若组数k=3，每组例数ni≤9，可查附表11，H界值表得出P值。 若组数k≥3，有ni≥10，则H近似服从? =k?1的?2分布。 Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 确定P值和作出推断结论 本例k=3，查附表12, 得P0.001。按? = 0.05水准拒绝H0，接受H1，差异有统计学意义，可认为不同程度再生障碍性贫血患者血清中可溶性CD8抗原水平有差别。 多组等级比较的检验假设 Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * H 的校正 当有相同秩次时，H 需校正： Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 　例8.3 某医院用三种复方小叶枇杷治疗老年性慢性支气管炎，数据见表8.2第(1)～(4)栏，试比较其疗效有无差异。 例题（page89） Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 例题（ page89 ） Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 建立检验假设 H0：三药疗效总体分布相同 ； H1：三药疗效总体分布不同或不全相同 ； ? = 0.05。 例题（ page89 ） Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 计算检验统计量H值 例题（ page89） Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics * 由于此资料的相同秩次很多，须校正： 例题 Evaluation only. Created with Aspose.Slides for .NET 3.5 Client Profile 5.2.0.0. Copyright 2004-2011 Aspose Pty Ltd. Biostatistics