Relaxation penalties and priors for plausible modeling of nonidentified bias sources
From MaRDI portal
Abstract: In designed experiments and surveys, known laws or design feat ures provide checks on the most relevant aspects of a model and identify the target parameters. In contrast, in most observational studies in the health and social sciences, the primary study data do not identify and may not even bound target parameters. Discrepancies between target and analogous identified parameters (biases) are then of paramount concern, which forces a major shift in modeling strategies. Conventional approaches are based on conditional testing of equality constraints, which correspond to implausible point-mass priors. When these constraints are not identified by available data, however, no such testing is possible. In response, implausible constraints can be relaxed into penalty functions derived from plausible prior distributions. The resulting models can be fit within familiar full or partial likelihood frameworks. The absence of identification renders all analyses part of a sensitivity analysis. In this view, results from single models are merely examples of what might be plausibly inferred. Nonetheless, just one plausible inference may suffice to demonstrate inherent limitations of the data. Points are illustrated with misclassified data from a study of sudden infant death syndrome. Extensions to confounding, selection bias and more complex data structures are outlined.
Recommendations
- Unbiased Bayes estimates and improper priors
- On intrinsic priors for nonnested models
- Bayesian penalty mixing: the case of a non-separable penalty
- Noninformative Priors and Nuisance Parameters
- Evaluating the tradeoff between bias and variance through use of prior probabilities
- Prior distributions from pseudo-likelihoods in the presence of nuisance parameters
- On hierarchical prior specifications and penalized likelihood
Cites work
- A General Approach to Analyzing Epidemiologic Data that Contain Misclassification Errors
- A New Perspective on Priors for Generalized Linear Models
- A note on estimating crude odds ratios in case-control studies with differentially misclassified exposure
- A note on partially Bayes inference and the linear model
- Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models
- Bayes Empirical Bayes
- Case-control analysis with partial knowledge of exposure misclassification probabilities
- Case-control studies: Assessing the effect of a confounding factor
- Choice as an alternative to control in observational studies. (With comments and a rejoinder).
- Common Structure of Smoothing Techniques in Statistics
- Error-free milestones in error prone measurements
- Estimating the Error Rates of Diagnostic Tests
- False Models and Post-Data Model Construction
- Families of distributions arising from distributions of order statistics
- Fusing multimodal biometrics with quality estimates via a Bayesian belief network
- Generalized conjugate priors for Bayesian analysis of risk and survival regressions
- How to Be a Better Bayesian
- scientific article; zbMATH DE number 3934086 (Why is no real title available?)
- scientific article; zbMATH DE number 47282 (Why is no real title available?)
- scientific article; zbMATH DE number 3519741 (Why is no real title available?)
- scientific article; zbMATH DE number 1349957 (Why is no real title available?)
- scientific article; zbMATH DE number 1059776 (Why is no real title available?)
- scientific article; zbMATH DE number 1487684 (Why is no real title available?)
- scientific article; zbMATH DE number 1834445 (Why is no real title available?)
- scientific article; zbMATH DE number 1911988 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- Identifiability, Improper Priors, and Gibbs Sampling for Generalized Linear Models
- Ignorance and uncertainty regions as inferential tools in a sensitivity analysis
- Incorporating prior beliefs about selection bias into the analysis of randomized trials with missing outcomes
- Interval estimation for messy observational data
- Maximum-likelihood and closed-form estimators of epidemiologic measures under misclassifi\-ca\-tion
- Measurement Error in Nonlinear Models
- Multiple-Bias Modelling for Analysis of Observational Data
- Observational studies.
- On model expansion, model contraction, identifiability and prior information: two illustrative scenarios involving mismeasured variables
- Sampling and Bayes' Inference in Scientific Modelling and Robustness
- Sensitivity Analysis for Incomplete Contingency Tables: The Slovenian Plebiscite Case
- The elements of statistical learning. Data mining, inference, and prediction
- The Impact of Prior Distributions for Uncontrolled Confounding and Response Bias
- The Performance of Random Coefficient Regression in Accounting for Residual Confounding
- When Should Epidemiologic Regressions Use Random Coefficients?
Cited in
(6)- The Impact of Prior Distributions for Uncontrolled Confounding and Response Bias
- Conflict diagnostics in directed acyclic graphs, with applications in Bayesian evidence synthesis
- A review of methods for research synthesis
- On model expansion, model contraction, identifiability and prior information: two illustrative scenarios involving mismeasured variables
- Comment: The need for syncretism in applied statistics
- Interval estimation for messy observational data
This page was built for publication: Relaxation penalties and priors for plausible modeling of nonidentified bias sources
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q903273)
