摘要
偏倚可分为选择偏倚、信息偏倚和混杂偏倚,从方向上又可分为正偏倚、负偏倚和颠倒偏倚。作为一种系统误差,偏倚可产生于兽医流行病学调查研究的设计、实施到推论等各个环节,可导致调查研究结论偏离乃至背离真实情况,必须加以控制。本文对兽医流行病学调查研究中可能出现的常见偏倚进行了阐述,并结合实例对不同的偏倚进行了分析,以期为广大兽医工作者更好地开展流行病学调查和疫病防控工作提供理论支持。
Biases,including selection biases,information biases and confounding biases,could be classified as positive,negative and switchover biases. As a systematic error,biases could be found in any sectors of veterinary epidemiological studies and should be avoided. In this paper,the common potential biases which may occur in veterinary epidemiological studies were analyzed with examples to provide support for veterinarians to conduct epidemiological investigation and disease control.
出处
《中国动物检疫》
CAS
2016年第4期67-71,共5页
China Animal Health Inspection
基金
科技部科技基础性专项(SQ2012FY3260033)
关键词
兽医流行病学研究
选择偏倚
信息偏倚
混杂偏倚
控制
veterinary epidemiological research
selection biases
information biases
confounding biases
control
