Bias is a systematic error that affects the observed measures of association between an outcome of interest and an exposure of interest in observational epidemiological studies. In contrast, effect modification is a true causal effect, where one exposure variable modifies the effect of another exposure variable on an outcome of interest. Eliminating bias and identifying effect modification are immensely important in observational studies. This chapter deals with (1) sources of bias (confounding, selection bias, and information bias), (2) selection bias and means to avoid it, (3) information bias and strategies to minimize it, (4) differential and nondifferential misclassification due to measurement error and its impact on measures of association, (5) concept of confounding, (6) ways to minimize confounding, (7) concept of effect modification, (8) synergistic and antagonistic effect modification, (9) quantitative and qualitative effect modification, and (10) how to identify and report effect modification using statistical analysis.
This is a preview of subscription content, log in via an institution to check access.
eBook EUR 85.59 Price includes VAT (France)
Hardcover Book EUR 105.49 Price includes VAT (France)
Tax calculation will be finalised at checkout
Purchases are for personal use only
