Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification
Abstract Background Log-binomial and robust (modified) Poisson regression models are popular approaches to estimate risk ratios for binary response variables. Previous studies have shown that comparatively they produce similar point estimates and standard errors. However, their performance under mod...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2018-06-01
|
| Series: | BMC Medical Research Methodology |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s12874-018-0519-5 |
| _version_ | 1818969927800848384 |
|---|---|
| author | Wansu Chen Lei Qian Jiaxiao Shi Meredith Franklin |
| author_facet | Wansu Chen Lei Qian Jiaxiao Shi Meredith Franklin |
| author_sort | Wansu Chen |
| collection | DOAJ |
| description | Abstract Background Log-binomial and robust (modified) Poisson regression models are popular approaches to estimate risk ratios for binary response variables. Previous studies have shown that comparatively they produce similar point estimates and standard errors. However, their performance under model misspecification is poorly understood. Methods In this simulation study, the statistical performance of the two models was compared when the log link function was misspecified or the response depended on predictors through a non-linear relationship (i.e. truncated response). Results Point estimates from log-binomial models were biased when the link function was misspecified or when the probability distribution of the response variable was truncated at the right tail. The percentage of truncated observations was positively associated with the presence of bias, and the bias was larger if the observations came from a population with a lower response rate given that the other parameters being examined were fixed. In contrast, point estimates from the robust Poisson models were unbiased. Conclusion Under model misspecification, the robust Poisson model was generally preferable because it provided unbiased estimates of risk ratios. |
| first_indexed | 2024-12-20T14:28:22Z |
| format | Article |
| id | doaj.art-3669b0e4339a453d8f22be1bb58c74ec |
| institution | Directory Open Access Journal |
| issn | 1471-2288 |
| language | English |
| last_indexed | 2024-12-20T14:28:22Z |
| publishDate | 2018-06-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Research Methodology |
| spelling | doaj.art-3669b0e4339a453d8f22be1bb58c74ec2022-12-21T19:37:42ZengBMCBMC Medical Research Methodology1471-22882018-06-0118111210.1186/s12874-018-0519-5Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecificationWansu Chen0Lei Qian1Jiaxiao Shi2Meredith Franklin3Kaiser Permanente Southern California, Department of Research and EvaluationKaiser Permanente Southern California, Department of Research and EvaluationKaiser Permanente Southern California, Department of Research and EvaluationDepartment of Preventive Medicine, Keck School of Medicine, University of Southern CaliforniaAbstract Background Log-binomial and robust (modified) Poisson regression models are popular approaches to estimate risk ratios for binary response variables. Previous studies have shown that comparatively they produce similar point estimates and standard errors. However, their performance under model misspecification is poorly understood. Methods In this simulation study, the statistical performance of the two models was compared when the log link function was misspecified or the response depended on predictors through a non-linear relationship (i.e. truncated response). Results Point estimates from log-binomial models were biased when the link function was misspecified or when the probability distribution of the response variable was truncated at the right tail. The percentage of truncated observations was positively associated with the presence of bias, and the bias was larger if the observations came from a population with a lower response rate given that the other parameters being examined were fixed. In contrast, point estimates from the robust Poisson models were unbiased. Conclusion Under model misspecification, the robust Poisson model was generally preferable because it provided unbiased estimates of risk ratios.http://link.springer.com/article/10.1186/s12874-018-0519-5Log-binomial regressionRobust (modified) Poisson regressionModel misspecificationRisk ratioLink function misspecification |
| spellingShingle | Wansu Chen Lei Qian Jiaxiao Shi Meredith Franklin Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification BMC Medical Research Methodology Log-binomial regression Robust (modified) Poisson regression Model misspecification Risk ratio Link function misspecification |
| title | Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification |
| title_full | Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification |
| title_fullStr | Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification |
| title_full_unstemmed | Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification |
| title_short | Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification |
| title_sort | comparing performance between log binomial and robust poisson regression models for estimating risk ratios under model misspecification |
| topic | Log-binomial regression Robust (modified) Poisson regression Model misspecification Risk ratio Link function misspecification |
| url | http://link.springer.com/article/10.1186/s12874-018-0519-5 |
| work_keys_str_mv | AT wansuchen comparingperformancebetweenlogbinomialandrobustpoissonregressionmodelsforestimatingriskratiosundermodelmisspecification AT leiqian comparingperformancebetweenlogbinomialandrobustpoissonregressionmodelsforestimatingriskratiosundermodelmisspecification AT jiaxiaoshi comparingperformancebetweenlogbinomialandrobustpoissonregressionmodelsforestimatingriskratiosundermodelmisspecification AT meredithfranklin comparingperformancebetweenlogbinomialandrobustpoissonregressionmodelsforestimatingriskratiosundermodelmisspecification |