医学卫生研究中的部分结果,可以用阳性或阴性反应表示,数据由清点呈阳性反应和阴性反应的例数而得。如病人治疗结果可归结为治愈、好转、无效、死亡等。类似这样的资料,其特点是先将观察单位按品质标志分组,再清点各组例数,这种资料称为计数资料。对于这类资料,可以用率、构成比等相对数进行统计描述。在对两个或多个率(构成比)作比较时,由于抽样误差的存在,研究者不能仅凭样本率(构成比)的大小直接下结论,而应该经过显著性检验。π2检验是计数资料分析中常用的统计方法之一。 Some results from medical and health research can be expressed as positive or negative reactions, with data obtained from the number of positive and negative reactions counted. If the patient treatment results can be attributed to cure, improvement, ineffective, death and so on. Similar to this information, which is characterized by the first observation units grouped by quality mark, and then count the number of cases in each group, this information is called count data. For such information, the availability rate, the composition of the relative number of statistical description. When comparing two or more rates (compositional ratios), because of the sampling error, researchers should not directly conclude based on the sample size (composition ratio) but should pass the significance test. π2 test is one of the commonly used statistical methods in counting data analysis.