论文部分内容阅读
前言非参数统计包含的内容很多,适用的范围较广泛,研究和应用发展也很快,其中秩和检验是比较好的一种。在医疗卫生工作中遇到一部分统计资料的总体是属于或己知属于某种理论分布(或经转换后使之成为标准理论分布),对这类数据应用“参数统计”是理想的统计分析方法。但有相当一部分搜集的数据其内在总体分布是不易辨别的,或者找不到某种合适的变量转换形式,或经转换也达不到某种理想的理论分布,这些数据若与假定的理论分布相差太远,参数统计就不适用。而秩和检验不受总体参数的制约,不要求样本一定来自特定的
Preface The non-parametric statistics contain many contents, the scope of application is more extensive, research and application development is also very fast, and the rank sum test is a better one. In the case of health care, some of the statistical data encountered are or belong to a certain theoretical distribution (or converted to a standard theoretical distribution). Applying “parameter statistics” to such data is an ideal statistical analysis method. . However, a considerable part of the collected data is not easily discernible in its overall internal distribution, or fails to find a suitable form of variable transformation, or fails to achieve an ideal theoretical distribution through transformation. If these data are related to the theoretical distribution of hypotheses The difference is too far away, parameter statistics do not apply. The rank sum test is not constrained by the overall parameters and does not require the sample to come from a specific