第八讲 流行病学研究中的偏差1 引言 流行病学的研究工作通常都以人群作为靶人群或参照人群(统计学称为总体)。我们希望通过试验人群(样本)中产生的研究结果代表总体的情况。但是,结果受到自身、被观察对象,以及研究方法等影响而产生偏差,使真实的联系被掩盖起来或造成虚假的联系。常见的偏差有抽样误差和偏性两大类。 Session Eight Deviations in Epidemiological Studies 1 Introduction Epidemiological research efforts generally target the population as a target or reference population (statistically referred to as population). We want to represent the overall picture of the research results generated in the experimental population (sample). However, the result is biased by itself, by the observed object, and by the method of research, so that the true linkages are obscured or cause false connections. Common deviations are sampling error and bias two categories.