Epidemiologic studies are beset by multiple biases and while sound study design and analysis may reduce many of these biases, some are likely to persist. Quantitative bias analysis is a set of techniques that allow researchers to adjust for a source of bias, after the study was conducted. Simple bias analysis techniques can be used to adjust for selection bias, misclassification and uncontrolled confounding. In this SERtalk, I will discuss simple bias analysis techniques as well as probabilistic bias analysis and Bayesian approaches that incorporate uncertainty in bias parameters as well.
