如前所述,分层和配对是控制混杂变量效应的基本方法。所谓分层(stratification),就是将病例一对照研究资料按有关变量的不同水平,分成若干层次,使得每层中的病例和对照数据直接可比。最简单的情况是将危险因素分为二个水平,即“暴露”与“未暴露”两类,对病例一对照数据进行分析。此时能借助通常的2×2表来概括数据: As mentioned earlier, stratification and pairing are the basic methods for controlling the effects of confounding variables. The so-called stratification (stratification) is to divide the case-control study data into levels according to different levels of relevant variables, so that the case and control data in each layer are directly comparable. In the simplest case, the risk factors are divided into two levels, “exposed” and “not exposed”, and the case-control data are analyzed. At this point the data can be summarized using the usual 2x2 tables: