Unmeasured bias in observational studies is often gauged by estimating effects that we think we know. Is a treatment associated with an outcome it should not affect? Is a treatment positively associated with an outcome for which a negative effect is anticipated? Information of this kind can inform a sensitivity analysis.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Known Effects in Observational Studies

  • Paul R. Rosenbaum

摘要

Unmeasured bias in observational studies is often gauged by estimating effects that we think we know. Is a treatment associated with an outcome it should not affect? Is a treatment positively associated with an outcome for which a negative effect is anticipated? Information of this kind can inform a sensitivity analysis.