Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias
A key objective of decomposition analysis is to identify a factor (the “mediator”) contributing to disparities in an outcome between social groups. In decomposition analysis, a scholarly interest often centers on estimating how much the disparity (e.g., health disparities between Black women and Whi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
De Gruyter
2023-03-01
|
| Series: | Journal of Causal Inference |
| Subjects: | |
| Online Access: | https://doi.org/10.1515/jci-2022-0031 |
| _version_ | 1797848709499715584 |
|---|---|
| author | Park Soojin Kang Suyeon Lee Chioun Ma Shujie |
| author_facet | Park Soojin Kang Suyeon Lee Chioun Ma Shujie |
| author_sort | Park Soojin |
| collection | DOAJ |
| description | A key objective of decomposition analysis is to identify a factor (the “mediator”) contributing to disparities in an outcome between social groups. In decomposition analysis, a scholarly interest often centers on estimating how much the disparity (e.g., health disparities between Black women and White men) would be reduced/remain if we set the mediator (e.g., education) distribution of one social group equal to another. However, causally identifying disparity reduction and remaining depends on the no omitted mediator–outcome confounding assumption, which is not empirically testable. Therefore, we propose a set of sensitivity analyses to assess the robustness of disparity reduction to possible unobserved confounding. We derived general bias formulas for disparity reduction, which can be used beyond a particular statistical model and do not require any functional assumptions. Moreover, the same bias formulas apply with unobserved confounding measured before and after the group status. On the basis of the formulas, we provide sensitivity analysis techniques based on regression coefficients and R2{R}^{2} values by extending the existing approaches. The R2{R}^{2}-based sensitivity analysis offers a straightforward interpretation of sensitivity parameters and a standard way to report the robustness of research findings. Although we introduce sensitivity analysis techniques in the context of decomposition analysis, they can be utilized in any mediation setting based on interventional indirect effects when the exposure is randomized (or conditionally ignorable given covariates). |
| first_indexed | 2024-04-09T18:32:00Z |
| format | Article |
| id | doaj.art-70e5795be74746faa2eb02dcf87fa41f |
| institution | Directory Open Access Journal |
| issn | 2193-3685 |
| language | English |
| last_indexed | 2024-04-09T18:32:00Z |
| publishDate | 2023-03-01 |
| publisher | De Gruyter |
| record_format | Article |
| series | Journal of Causal Inference |
| spelling | doaj.art-70e5795be74746faa2eb02dcf87fa41f2023-04-11T17:07:16ZengDe GruyterJournal of Causal Inference2193-36852023-03-011111555910.1515/jci-2022-0031Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable biasPark Soojin0Kang Suyeon1Lee Chioun2Ma Shujie3School of Education, University of California, Riverside, California, United States of AmericaDepartment of Statistics, University of California, Riverside, California, United States of AmericaDepartment of Sociology, University of California, Riverside, California, United States of AmericaDepartment of Statistics, University of California, Riverside, California, United States of AmericaA key objective of decomposition analysis is to identify a factor (the “mediator”) contributing to disparities in an outcome between social groups. In decomposition analysis, a scholarly interest often centers on estimating how much the disparity (e.g., health disparities between Black women and White men) would be reduced/remain if we set the mediator (e.g., education) distribution of one social group equal to another. However, causally identifying disparity reduction and remaining depends on the no omitted mediator–outcome confounding assumption, which is not empirically testable. Therefore, we propose a set of sensitivity analyses to assess the robustness of disparity reduction to possible unobserved confounding. We derived general bias formulas for disparity reduction, which can be used beyond a particular statistical model and do not require any functional assumptions. Moreover, the same bias formulas apply with unobserved confounding measured before and after the group status. On the basis of the formulas, we provide sensitivity analysis techniques based on regression coefficients and R2{R}^{2} values by extending the existing approaches. The R2{R}^{2}-based sensitivity analysis offers a straightforward interpretation of sensitivity parameters and a standard way to report the robustness of research findings. Although we introduce sensitivity analysis techniques in the context of decomposition analysis, they can be utilized in any mediation setting based on interventional indirect effects when the exposure is randomized (or conditionally ignorable given covariates).https://doi.org/10.1515/jci-2022-0031interventional indirect effectunobserved confoundingdisparity reductiondisparity remainingrobustness value62d20 |
| spellingShingle | Park Soojin Kang Suyeon Lee Chioun Ma Shujie Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias Journal of Causal Inference interventional indirect effect unobserved confounding disparity reduction disparity remaining robustness value 62d20 |
| title | Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias |
| title_full | Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias |
| title_fullStr | Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias |
| title_full_unstemmed | Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias |
| title_short | Sensitivity analysis for causal decomposition analysis: Assessing robustness toward omitted variable bias |
| title_sort | sensitivity analysis for causal decomposition analysis assessing robustness toward omitted variable bias |
| topic | interventional indirect effect unobserved confounding disparity reduction disparity remaining robustness value 62d20 |
| url | https://doi.org/10.1515/jci-2022-0031 |
| work_keys_str_mv | AT parksoojin sensitivityanalysisforcausaldecompositionanalysisassessingrobustnesstowardomittedvariablebias AT kangsuyeon sensitivityanalysisforcausaldecompositionanalysisassessingrobustnesstowardomittedvariablebias AT leechioun sensitivityanalysisforcausaldecompositionanalysisassessingrobustnesstowardomittedvariablebias AT mashujie sensitivityanalysisforcausaldecompositionanalysisassessingrobustnesstowardomittedvariablebias |