Adjust for Selection Bias [9] [12]

Include the variable associated with selection in the analysis, to reduce selection bias in a similar way to confounding [13] .

Bias Breaking [5]

A method which produces bias-adjusted estimates for the odds ratios in case control studies.

Imputation [8]

Often multiple imputation; replace missing values with reasonable estimates using the collected data.

Population Data [7]

Use population data in place of control data in a case control study.

Post-Stratification [14]

Classify unmatched samples of cases and controls based on their values on one or more of the variables in the study. Similar to stratified sampling or frequency matching.

Predict the Bias [15] - [17]

Use information from non-participants to try to predict the amount of bias present.

Propensity Score [10]

Can be used to match cases and controls, or as an additional covariate during analysis.

Sensitivity Analysis [6] [12]

A method for estimating the direction and magnitude of the bias.

Stratification [11]

Calculate estimates conditional on at least one other variable, which can lead to unbiased estimates within strata.

Weighting [4]

Usually inverse probability weighting; use external data to assign each subject a weight which is the inverse of their probability of selection, to allow them to represent non-participants.