Bias and Sensitivity Analyses for Linear Front-Door Models
The front-door model allows unbiased estimation of a total effect in the presence of unobserved confounding. This guarantee of unbiasedness hinges on a set of assumptions that can be violated in practice. We derive formulas that quantify the amount of bias for specific violations, and contrast them...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PsychOpen GOLD/ Leibniz Institute for Psychology
2023-09-01
|
| Series: | Methodology |
| Subjects: | |
| Online Access: | https://doi.org/10.5964/meth.9205 |
| _version_ | 1797320629797519360 |
|---|---|
| author | Felix Thoemmes Yongnam Kim |
| author_facet | Felix Thoemmes Yongnam Kim |
| author_sort | Felix Thoemmes |
| collection | DOAJ |
| description | The front-door model allows unbiased estimation of a total effect in the presence of unobserved confounding. This guarantee of unbiasedness hinges on a set of assumptions that can be violated in practice. We derive formulas that quantify the amount of bias for specific violations, and contrast them with bias that would be realized from a naive estimator of the effect. Some violations result in simple, monotonic increases in bias, while others lead to more complex bias, consisting of confounding bias, collider bias, and bias amplification. In some instances, these sources of bias can (partially) cancel each other out. We present ways to conduct sensitivity analyses for all violations, and provide code that performs sensitivity analyses for the linear front-door model. We finish with an applied example of the effect of math self-efficacy on educational achievement. |
| first_indexed | 2024-03-08T04:45:47Z |
| format | Article |
| id | doaj.art-699322e6e9e74c2ea442a6f84aff5a28 |
| institution | Directory Open Access Journal |
| issn | 1614-2241 |
| language | English |
| last_indexed | 2024-03-08T04:45:47Z |
| publishDate | 2023-09-01 |
| publisher | PsychOpen GOLD/ Leibniz Institute for Psychology |
| record_format | Article |
| series | Methodology |
| spelling | doaj.art-699322e6e9e74c2ea442a6f84aff5a282024-02-08T10:51:08ZengPsychOpen GOLD/ Leibniz Institute for PsychologyMethodology1614-22412023-09-0119325628210.5964/meth.9205meth.9205Bias and Sensitivity Analyses for Linear Front-Door ModelsFelix Thoemmes0https://orcid.org/0000-0001-5689-2659Yongnam Kim1Cornell University, Ithaca, NY, USASeoul National University, Seoul, South KoreaThe front-door model allows unbiased estimation of a total effect in the presence of unobserved confounding. This guarantee of unbiasedness hinges on a set of assumptions that can be violated in practice. We derive formulas that quantify the amount of bias for specific violations, and contrast them with bias that would be realized from a naive estimator of the effect. Some violations result in simple, monotonic increases in bias, while others lead to more complex bias, consisting of confounding bias, collider bias, and bias amplification. In some instances, these sources of bias can (partially) cancel each other out. We present ways to conduct sensitivity analyses for all violations, and provide code that performs sensitivity analyses for the linear front-door model. We finish with an applied example of the effect of math self-efficacy on educational achievement.https://doi.org/10.5964/meth.9205causal inferencefront-doorbiasmeasurement errorsensitivity analysis |
| spellingShingle | Felix Thoemmes Yongnam Kim Bias and Sensitivity Analyses for Linear Front-Door Models Methodology causal inference front-door bias measurement error sensitivity analysis |
| title | Bias and Sensitivity Analyses for Linear Front-Door Models |
| title_full | Bias and Sensitivity Analyses for Linear Front-Door Models |
| title_fullStr | Bias and Sensitivity Analyses for Linear Front-Door Models |
| title_full_unstemmed | Bias and Sensitivity Analyses for Linear Front-Door Models |
| title_short | Bias and Sensitivity Analyses for Linear Front-Door Models |
| title_sort | bias and sensitivity analyses for linear front door models |
| topic | causal inference front-door bias measurement error sensitivity analysis |
| url | https://doi.org/10.5964/meth.9205 |
| work_keys_str_mv | AT felixthoemmes biasandsensitivityanalysesforlinearfrontdoormodels AT yongnamkim biasandsensitivityanalysesforlinearfrontdoormodels |