The prior can generally only be understood in the context of the likelihood

Abstract

A key sticking point of Bayesian analysis is the choice of prior distribution, and there is a vast literature on potential defaults including uniform priors, Jeffreys' priors, reference priors, maximum entropy priors, and weakly informative priors. These methods, however, often manifest a key conceptual tension in prior modeling: a model encoding true prior information should be chosen without reference to the model of the measurement process, but almost all common prior modeling techniques are implicitly motivated by a reference likelihood. In this paper we resolve this apparent paradox by placing the choice of prior into the context of the entire Bayesian analysis, from inference to prediction to model evaluation.

13 pages

References in corpus (2)

• A weakly informative default prior distribution for logistic and other regression models
• Adaptive Bayesian estimation using a Gaussian random field with inverse Gamma bandwidth

Cited by in corpus (30)

• Social physics
• Calibrate, Emulate, Sample
• Detecting and diagnosing prior and likelihood sensitivity with power-scaling
• Model-independent constraints on ${\Omega }_m$ and $H(z)$ from the link between geometry and growth
• The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learning
• Applying Bayesian Analysis Guidelines to Empirical Software Engineering Data: The Case of Programming Languages and Code Quality
• Poroelasticity as a Model of Soft Tissue Structure: Hydraulic Permeability Inference for Magnetic Resonance Elastography in Silico
• Sparsity-Exploiting Anchor Placement for Localization in Sensor Networks
• Not All Requirements Prioritization Criteria Are Equal at All Times: A Quantitative Analysis
• The Risks of Ranking: Revisiting Graphical Perception to Model Individual Differences in Visualization Performance
• Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
• A guide to state-space modeling of ecological time series
• Hyperparameter Optimization for Randomized Algorithms: A Case Study on Random Features
• exoALMA. VIII. Probabilistic Moment Maps and Data Products using Non-parametric Linear Models
• Opaque prior distributions in Bayesian latent variable models
• Bayesian models are better than frequentist models in identifying differences in small datasets comprising phonetic data
• Efficient estimation and correction of selection-induced bias with order statistics
• Informative Bayesian model selection for RR Lyrae star classifiers
• Had enough of experts? Quantitative knowledge retrieval from large language models
• Knowing what to know: Implications of the choice of prior distribution on the behavior of adaptive design optimization
• Atmospheric characterization of hot Jupiters using hierarchical models of Spitzer observations
• Modern Statistical Models and Methods for Estimating Fatigue-Life and Fatigue-Strength Distributions from Experimental Data
• Six Maxims of Statistical Acumen for Astronomical Data Analysis
• Modeling racial/ethnic differences in COVID-19 incidence with covariates subject to non-random missingness
• Design and Structure Dependent Priors for Scale Parameters in Latent Gaussian Models
• Prediction can be safely used as a proxy for explanation in causally consistent Bayesian generalized linear models
• The 2025 Evaluation of Experimental Thermonuclear Reaction Rates (ETR25)
• Gaussian distributional structural equation models: A framework for modeling latent heteroscedasticity
• Bayesian generalized method of moments applied to pseudo-observations in survival analysis
• $\text{Exoformer}$: Accelerating Bayesian atmospheric retrievals with transformer neural networks