Confounders and Using the Method of Propensity Scores
Chin Lin, Rui Imamura, Felipe Fregni
Abstract
Chin Lin, Rui Imamura, Felipe Fregni
Abstract
This chapter explores the important issue of confounding in observational studies. The potential imbalances that result for not controlling assignment of treatment or exposure may lead to imbalance of variables that are associated with both treatment and intervention (or exposure) thus confounding results. Therefore, in this context, a potential relationship between an intervention and an outcome could be invalid. This chapter therefore explains basic definitions of confounding and presents some methods to control for confounders, highlighting the use of the propensity score, which is considered a robust method for this purpose. Different techniques of adjustment using propensity score (matching, stratification, regression, and weighting) are also discussed. This chapter concludes with a case discussion about confounding and how to address it.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This chapter explores the important issue of confounding in observational studies. The potential imbalances that result for not controlling assignment of treatment or exposure may lead to imbalance of variables that are associated with both treatment and intervention (or exposure) thus confounding results. Therefore, in this context, a potential relationship between an intervention and an outcome could be invalid. This chapter therefore explains basic definitions of confounding and presents some methods to control for confounders, highlighting the use of the propensity score, which is considered a robust method for this purpose. Different techniques of adjustment using propensity score (matching, stratification, regression, and weighting) are also discussed. This chapter concludes with a case discussion about confounding and how to address it.
Key concepts: Propensity score matching, Confounding, Observational study, Context (archaeology), Matching (statistics), Weighting, Outcome (game theory), Medicine