STAD Seminar: Selection bias for marginal parameters

Abstract: Selection bias is a systematic error that can occur when subjects are included or excluded in the analysis based on some selection criteria for the study population. This type of bias can jeopardize the validity of the study, not only in the case of generalization to the total population but also for the selected study population. Most previous studies on selection bias ignore both confounders and sampling variability and suggested approaches consider selection bias within strata of covariates. Instead of stratum-specific conditional analysis, we here consider selection bias for marginal parameters. The selection bias for the marginal parameters depends on two things, the magnitude of selection bias within the strata and the probability distribution of the pre-treatment covariates. This is ongoing work, and we aim to apply our results to a numerical example studying the effect of pregnancy on long-term renal complications among type 1 diabetic women.

This is a seminar in our seminar series on Statistics, AI and Data Science (STAD).