论文部分内容阅读
在病例—对照研究中,有些未被控制的因素在一般x~2检验中,有可能掩盖研究因素与结果的真实联系性,而成为结果判定的混杂因素。在调查设计阶段未被控制的因素,在资料分析阶段有些混杂因素可通过适当的分析方法加以控制。其方法之一就是分层分析(Mantel-Haenszel分析方法)。其原理是将混杂因素同等分布于病例组和对照组,使之对病例和对照产生相同的影响,然后用分层x~2 M-Ⅱ检验。在分析因素不太多的情况下,此方法是较为便利的,它适用于有无混杂因素的两种情况。
In the case-control study, some uncontrolled factors in the general x~2 test may obscure the real connection between research factors and results, and become confounding factors in the outcome determination. Factors that are not controlled during the survey design phase may be confounded by appropriate analysis methods during the data analysis phase. One of the methods is hierarchical analysis (Mantel-Haenszel analysis method). The principle is to distribute the confounding factors equally in the case group and the control group so that they have the same effect on the case and the control, and then use the stratified x~2 M-II test. This method is more convenient when there are not many analysis factors, and it is applicable to both cases with or without confounding factors.