论文部分内容阅读
临床科研的各阶段都可产生偏倚,大致可分为3大类:(1)选择性偏倚。产生于临床科研的设计阶段,由于选择研究对象的方法不正确所造成。严格的科研设计,强调随机化原则进行分组常能有效地防止选择性偏倚。(2)测量性偏倚。产生于科研的实施阶段,系由于2组病人所采用的测量和观察方法不一致所造成的系统误差。采用盲法测量能有效地防止测量性偏倚。(3)混杂性偏倚。在评价被研究的因素和疾病之间的关系时,如果存在外来因素,这因素与该病和研
Biases can be generated in all stages of clinical research and can be broadly categorized into three broad categories: (1) Selective biases. Produced in the design phase of clinical research, due to the method of selecting the object of study is not correct. Strict scientific research design, emphasizing the principle of randomization can often be effective in preventing selective bias. (2) Measurement bias. Produced in the implementation phase of scientific research, due to the two groups of patients used by the measurement and observation of the inconsistencies caused by systematic errors. Blind measurement can effectively prevent measurement bias. (3) Mixed bias. In assessing the relationship between the factors studied and the disease, if there is a foreign factor, this factor is associated with the disease and research