错误分组的发生是由于使用了接触代用指标。错误分组表现出比实际情况为小的相对危险,同时,各统计检验的效能有所降低。在流行病学调查中,调查对象的错误分组,既能使效应指标的估计值产生偏倚,又能使统计检验效能受影响。当一个调查对象被放到错误的“接触”水平分组中,或放到错误的“结果”分组中,或二类分组都放错了,就造成 The occurrence of false groupings is due to the use of contact surrogate indicators. Wrong subgroups show relative risks that are smaller than they actually are, and the effectiveness of the statistical tests is somewhat reduced. In the epidemiological survey, the incorrect grouping of the respondents can not only bias the estimation of the effect index but also affect the performance of the statistical test. When a respondent is placed in the wrong “touch” level group, or in the wrong “result” group, or both are misplaced, this results in