Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper)
Open accessThe DOI matches its Crossref record; 42 of the 42 references it cites were found in a registry or an index; 1 of the 23 sources its claims cite were checked for support.
Abstract
Various noninformative prior distributions have been suggested for scale parameters in hierarchical models. We construct a new folded-noncentral-$t$ family of conditionally conjugate priors for hierarchical standard deviation parameters, and then consider noninformative and weakly informative priors in this family. We use an example to illustrate serious problems with the inverse-gamma family of "noninformative" prior distributions. We suggest instead to use a uniform prior on the hierarchical standard deviation, using the half-$t$ family when the number of groups is small and in other settings where a weakly informative prior is desired. We also illustrate the use of the half-$t$ family for hierarchical modeling of multiple variance parameters such as arise in the analysis of variance.
Overview
CoChat Summary
TL;DR
Gelman argues that inverse-gamma “noninformative” priors for hierarchical variance parameters can badly distort posterior inference, especially when the group-level variance is near zero or the number of groups is small. He recommends a uniform prior on the hierarchical standard deviation when groups are not few, and half-t/half-Cauchy weakly informative priors otherwise, developed within a folded-noncentral-t conditionally conjugate family.
Key findings
- The inverse-gamma(ε, ε) family is not reliably noninformative: as ε → 0 it yields an improper posterior, and in the 8-schools example inverse-gamma(1, 1) and inverse-gamma(0.001, 0.001) priors on σ²_α constrained posterior inferences and increased shrinkage relative to a uniform prior on σ_α.
- A uniform prior on σ_α performed well for the 8-schools data (J = 8), but with only J = 3 schools it produced an extremely long right tail and undershrinkage; a half-Cauchy prior with scale A = 25 removed much of that unrealistic upper tail.
- For multiple variance components, Gelman recommends a hierarchical half-Cauchy model with scale estimated from data; in the 5 × 5 × 2 split-plot latin square Anova with nine variance components, the Cauchy scale parameter A was estimated at 1.8 with a 95% posterior interval of [0.5, 5.1].
- The folded-noncentral-t family, obtained via a redundant multiplicative parameterization, provides conditionally conjugate priors for σ_α and includes the improper uniform density (ν = −1) and proper half-Cauchy (ν = 1) as special cases.
Methods
- Conceptual/comment paper using a simple two-level normal hierarchical model: y_ij ∼ N(μ + α_j, σ²_y), α_j ∼ N(0, σ²_α), focusing on the prior for σ_α.
- Illustrates with the 8-schools example (J = 8, 6000 Bugs simulation draws), a 3-schools subset, and a 5 × 5 × 2 split-plot latin square Anova with nine variance components.
- Uses a parameter-expanded form y_ij ∼ N(μ + ξη_j, σ²_y), η_j ∼ N(0, σ²_η), with σ_α = |ξ|σ_η, to construct and compare uniform, inverse-gamma, half-Cauchy, and hierarchical half-Cauchy priors.
Limitations
- The paper states that noninformative and weakly informative priors are inherently provisional: after fitting, one should inspect the posterior and revise the prior if it does not make sense.
- The paper notes that the uniform prior on σ_α has a positive miscalibration and yields an improper posterior for J = 1 or 2; inverse-gamma inferences are sensitive to ε; and half-Cauchy scale choices are weakly informative rather than fully objective.
- Evident from the text: the recommendations rest on illustrative examples and conceptual argument rather than a broad simulation study or systematic empirical comparison, and extensions to nonlinear, multivariate, and more complex hierarchical models are left for further work.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 1 Not valid · 14 Not checked · 56 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Similarly, there are no always-nonnegative classical unbiased estimators of σ α or σ 2 α in the hierarchical model.”
- Also stated in: § 7.1 Prior distributions for variance parameters · plain
“Similarly, there are no alwaysnonnegative classical unbiased estimators of σ α or σ 2 α in the hierarchical model.”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 2.4 Calibration
What the sources say
- Could not verifycites Bickel and Blackwell (1967) matched to “A Note on Bayes Estimates” (1967) sourceExperimental — accuracy not yet measured
The introduction (p. 1) and Theorem 1.3 (p. 3) concern when an unbiased Bayes estimate must have zero Bayes risk. Searches for hierarchical models, σₐ, σ²ₐ, and always-nonnegative classical unbiased estimators found no passage addressing the claimed nonexistence result. (LLM, conf=0.77)
- Not checked — no full text of the sourcecites Meng and Zaslavsky (2002) matched to “Single observation unbiased priors” (2002) source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 7.1 Prior distributions for variance parameters · plain
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If J is low, the uniform prior density tends to lead to high estimates of σ α , as discussed in Section 5.2.”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 7.1 Prior distributions for variance parameters
hedgedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In contrast, the hierarchical half-Cauchy model performs a great deal of shrinkage, especially of the high ranges of the intervals.”
- Also stated in: § Bugs model: a half-Cauchy prior distribution on sigma.theta is induced # using a normal prior on xi and an inverse-gamma on tau.eta model { for (j in 1:J){ # J = the number of schools y[j] ~dnorm (theta[j], tau.y[j]) # data model: the likelihood theta[j] <-mu.theta+ xi*eta[j] tau.y[j] <-pow(sigma.y[j],-2) } xi ~dnorm (0, tau.xi) tau.xi <-pow(prior.scale,-2) for (j in 1:J){ eta[j] ~dnorm (0, tau.eta) # hierarchical model for theta } tau.eta ~dgamma (.5, .5)# chi^2 with 1 d.f.sigma.theta<-abs(xi)/sqrt(tau.eta)# cauchy = normal/sqrt(chi^2) mu.theta ~dnorm (0.0, 1.0E-6) # noninformative prior on mu } · plain
“The half-Cauchy model gives much sharper inferences, using the partial pooling that comes with fitting a hierarchical model.”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 6.2 Superpopulation and finite-population standard deviations
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Bugs model: a half-Cauchy prior distribution on sigma.theta is induced # using a normal prior on xi and an inverse-gamma on tau.eta model { for (j in 1:J){ # J = the number of schools y[j] ~dnorm (theta[j], tau.y[j]) # data model: the likelihood theta[j] <-mu.theta+ xi*eta[j] tau.y[j] <-pow(sigma.y[j],-2) } xi ~dnorm (0, tau.xi) tau.xi <-pow(prior.scale,-2) for (j in 1:J){ eta[j] ~dnorm (0, tau.eta) # hierarchical model for theta } tau.eta ~dgamma (.5, .5)# chi^2 with 1 d.f.sigma.theta<-abs(xi)/sqrt(tau.eta)# cauchy = normal/sqrt(chi^2) mu.theta ~dnorm (0.0, 1.0E-6) # noninformative prior on mu } · plain
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This prior distribution is even more sharply peaked near zero and further distorts posterior inferences, with the problem arising because the marginal likelihood for σ α remains high near zero.”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 5.1 Noninformative prior distributions for the 8-schools problem
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The uniform prior distribution seems fine for the 8-school analysis”
- Also stated in: § 5.2 Weakly informative prior distribution for the 3-schools problem · plain
“the default uniform prior gave reasonable results”
- Also stated in: § 7.1 Prior distributions for variance parameters · hedged
“Figure 1 illustrates the generally robust properties of the uniform prior density on σ α .”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 5.2 Weakly informative prior distribution for the 3-schools problem
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 5.2 Weakly informative prior distribution for the 3-schools problem · plain
26 more claims
Create a free account to see all 71 claims and every source verdict.
- It may state two separate claims as one. 34
- Probably not a substantial claim. 12
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Because the uniform distribution tends to unduly favor presence of heterogeneity in discrimination and calibration estimates across studies,58,59”
From A framework for meta-analysis of prediction model studies with binary and time-to-event outcomes (2019) · cites this paper as [59]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 7.1 Prior distributions for variance parameters
“we see conditional conjugacy in the wider class of half-t distributions on σ α , which include the uniform and half-Cauchy densities on σ α (as well as inverse-gamma on σ 2 α ) as special cases.”
§ Prior distributions for variance parameters in hierarchical models(Comment on Article by Browne and Draper) > 3.2 Folded-noncentral-t prior distribution for σ α
“In the limit that the denominator is specified exactly, we have a folded normal distribution; conversely, specifying the numerator exactly yields the square-root-inverse-χ 2 distribution for σ α , as in Section 3.1.”
Citations
Cites
Gelman, A., Carlin, J. B., Stern, H. S., & Rubin, D. B. (2003) · doi:10.1201/9780429258480 · Cited by 27,276
Self-citationKass, R. E., & Raftery, A. E. (1995) · Journal of the American Statistical Association · doi:10.1080/01621459.1995.10476572 · Cited by 12,558
Sampling-based approaches to calculating marginal densities
Gelfand, A. E., & Smith, A. F. M. (1990) · Journal of the American Statistical Association · doi:10.1080/01621459.1990.10476213 · Cited by 6,606 · cited 1×
“Typically, enough data will be available to estimate µ and σy that one can use any reasonable noninformative prior distribution—for example, p(µ, σy) ∝ 1 or p(µ, log σy) ∝ 1.”
Introducing Multilevel Modeling
Kreft, I., & de Leeuw, J. (1998) · doi:10.4135/9781849209366 · Cited by 2,997
POSTERIOR PREDICTIVE ASSESSMENT OF MODEL FITNESS VIA REALIZED DISCREPANCIES
Gelman, A., Meng, X., & Stern, H. S. (1996) · Cited by 2,638
Self-citationBayes and Empirical Bayes Methods for Data Analysis
(2001) · Technometrics · doi:10.1198/tech.2001.s608 · Cited by 2,098
INADMISSIBILITY OF THE USUAL ESTIMATOR FOR THE MEAN OF A MULTIVARIATE NORMAL DISTRIBUTION
Stein, C. (1956) · Contribution to the Theory of Statistics · doi:10.1525/9780520313880-018 · Cited by 1,841
Bayesian Approaches to Clinical Trials and Health‐Care Evaluation
Spiegelhalter, D. J., Abrams, K. R., & Myles, J. P. (2003) · Wiley eBooks · doi:10.1002/0470092602 · Cited by 1,673
Parametric empirical Bayes inference: theory and applications
Morris, C. N. (1983) · Journal of the American Statistical Association · doi:10.1080/01621459.1983.10477920 · Cited by 1,448 · cited 1×
“On the theoretical side, hierarchical models allow a more “objective” approach to inference by estimating the parameters of prior distributions from data rather than requiring them to be specified using subjective information (see James and Stein, 1960, Efron and Morris, 1975, and Morris, 1983).”
The Selection of Prior Distributions by Formal Rules
Kass, R. E., & Wasserman, L. (1996) · Journal of the American Statistical Association · doi:10.1080/01621459.1996.10477003 · Cited by 1,308
Research works printed but not linked: 5
- Box, G. E. P. and Tiao, G. C. (1973). Bayesian Inference in Statistical Analysis. Reading, Mass.: Addison-Wesley.
- James, W. and Stein, C. (1960). "Estimation with quadratic loss." In Neyman, J. (ed.), Proceedings of the Fourth Berkeley Symposium, volume 1, 361–380. Berkeley: University of California Press.
- Johnson, N. L. and Kotz, S. (1972). Distributions in Statistics. New York: Wiley. 4 vols.
- Liu, C. (2001). "Bayesian analysis of multivariate probit models. Discussion of “The art of data augmentation” by D. A. van Dyk and X. L. Meng." Journal of Computational and Graphical Statistics, 10: 75–81.
- Raudenbush, S. W. and Bryk, A. S. (2002). Hierarchical Linear Models. Thousand Oaks, Calif.: Sage., second edition.
Cited by
Items 3,701–3,800 of 4,089
Li, G., & An, L. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/2726602
Supplementary material to "Functional ANOVA for Carbon Flux Estimates from Remote Sensing Data"
Hobbs, J., Katzfuß, M., Nguyen, H., et al. (2022) · doi:10.5194/gmd-2022-230-supplement
Gomer, B., & Yuan, K. (2022) · Elsevier eBooks · doi:10.1016/b978-0-12-818630-5.10090-9 · Cited by 2
Correction to our Reply on RC2
Ringel, L. M. (2022) · doi:10.5194/hess-2022-209-ac3
Kavelaars, X., Mulder, J., & Kaptein, M. C. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2212.03473
(2022) · Climate Change and Law Collection · doi:10.1163/9789004322714_cclc_2020-0148-0747
Thomas, S., & Tu, W. (2022) · Wiley StatsRef: Statistics Reference Online · doi:10.1002/9781118445112.stat08392
Calvo, G., Armero, C., Gómez‐Rubio, V., et al. (2022) · Springer proceedings in mathematics & statistics · doi:10.1007/978-3-031-16427-9_11
Evaluation of the Mean Control Chart Under a Bayesian Approach
Guevara, I. C. R., & González-Álvarez, N. (2022) · Revista Colombiana de Estadística · doi:10.15446/rce.v45n1.93588
“• Hyperparameters of the prior distribution of σ−2: following Gelman (2006), we take ν02 = ν0s 2 0 2 = ν, with ν = 0.001 and 1, where large values indicate little uncertainty regarding to σ−2.”
MethodInfluentialA Bayesian model for identifying cancer subtypes from paired methylation profiles
Fan, Y., Chan, A., Zhu, J., Leung, S. Y., & Fan, X. (2022) · Briefings in Bioinformatics · doi:10.1093/bib/bbac568
Guillaume, B., Boukbida, H. A., Bakker, G., Bieganowski, A., Brostaux, Y., Cornelis, W., Durner, W., Hartmann, C., Iversen, B. V., Javaux, M., et al. (2023) · doi:10.5194/egusphere-2022-1496-supplement
A Bayesian approach to data-driven discovery of nonlinear dynamic equations
North, J. S. (2022) · MOspace Institutional Repository (University of Missouri) · doi:10.32469/10355/93981
Liu, P., Li, C., He, Z., et al. (2022) · International Conference on Electrical Engineering and Control Technologies (CEECT) · doi:10.1109/ceect55960.2022.10030634
“The reason to use this design to use the prior distribution to represent the model parameters is following the work of [14], [15].”
MethodAnjoy, P. (2023) · Journal of Quantitative Economics · doi:10.1007/s40953-023-00338-x
Rojas-Mora, J. M., & Guevara, I. R. (2022) · CIENCIA EN DESARROLLO · doi:10.19053/01217488.v13.n1.2022.13655
Modelling admission lengths within psychiatric intensive care units
Dye, S., Sethi, F., Kearney, T., Rose, E., Penfold, L., Campbell, M., & Valsraj, K. (2023) · BMJ Health & Care Informatics · doi:10.1136/bmjhci-2022-100685
Bayesian Analysis of Multivariate Matched Proportions with Sparse Response
Meyer, M. J., Cheng, H., & Knutson, K. H. (2023) · Statistics in Biosciences · doi:10.1007/s12561-023-09368-8
Pan, H., & Yuan, Y. (2023) · Springer eBooks · doi:10.1007/978-981-19-8176-0_8
A Bayesian workflow for the analysis and reporting of international large-scale surveys
Kaplan, D. E., & Harra, K. (2023) · OECD education working papers · doi:10.1787/588c4a12-en
Probability Distributions and Generating Functions
Wakefield, J. (2012) · Springer series in statistics · doi:10.1007/978-1-4419-0925-1_16
Improving Predictions of Response Propensities for Effective Adaptive Survey Design (ASD)
Wu, S. (2023) · Utrecht University Repository (Utrecht University) · doi:10.33540/1791
Kundu, D., Mitra, R., Albert, P. S., & Gaskins, J. (2023) · Statistics in Medicine · doi:10.1002/sim.9809
Advancing the Methodology for Estimating the Effects of Safety Treatments
Anarkooli, A. J. (2023) · doi:10.32920/23541900
Advancing the Methodology for Estimating the Effects of Safety Treatments
Anarkooli, A. J. (2023) · doi:10.32920/23541900.v1
Discussion of: Specifying prior distributions in reliability applications
Arnold, R. (2023) · Applied Stochastic Models in Business and Industry · doi:10.1002/asmb.2791
A Goodness of Fit Test for a Survival and Count Bayesian Joint Model: In the Presence of Clusters
Kumaranathunga, K. U. S., & Sooriyarachchi, M. R. (2023) · Sri Lankan Journal of Applied Statistics · doi:10.4038/sljastats.v24i1.8090
The association between socioeconomic status and C-reactive protein in Bayesian perspective
Procopio, A., & Samuel, R. (2023) · SSM: Population Health · doi:10.1016/j.ssmph.2023.101464
“…β0 ) = N ( μy, σy ) p ( σβi ) = C ( 0, σy ) p ⎛ ⎜ ⎝βj ⎞ ⎟ ⎠ = N ⎛ ⎜ ⎝0, 1 ⎞ ⎟ ⎠p ⎛ ⎜ ⎝βi ⎞ ⎟ ⎠ = N ⎛ ⎜ ⎝0, 1 ⎞ ⎟ ⎠p ⎛ ⎜ ⎝σ ⎞ ⎟ ⎠ = N ⎛ ⎜ ⎝0, σy ⎞ ⎟ ⎠p ⎛ ⎜ ⎝ν ⎞ ⎟ ⎠ = e ⎛ ⎜ ⎝1 / 29 ⎞ ⎟ ⎠ (2) The prior distributions of Model 1 and Model 2 follow the suggestions from Gelman (2006) and Kruschke (2014).”
Bayesian Learning of Personalized Longitudinal Biomarker Trajectory
Zhou, S., Huang, X., Shen, C., & Kantarjian, H. M. (2023) · Annals of Data Science · doi:10.1007/s40745-023-00486-0
Longitudinal binary response models using alternative links for medical data
Huayanay, A. D. L. C., Bazán, J. L., & Diniz, C. A. R. (2023) · Brazilian Journal of Probability and Statistics · doi:10.1214/23-bjps572
Man, W., Wei, X., & Sun, J. (2023) · International Conference on Geographic Information and Remote Sensing Technology (GIRST) · doi:10.1117/12.3007396
“All random effects (including σε and σv ) priors follow a half-normal distribution N+∞(0,10) as suggested by Andrew Gelman for Browne and Draper[14].”
MethodResolving the climatic and ecological drivers of geographical gradients in avian sexual selectionPreprint
Barber, R. A., Yang, J., Yang, C., et al. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.09.01.555923
A Bayesian growth mixture model for complex survey data: Clustering postdisaster PTSD trajectories
Anthopolos, R., Chen, Q., Sedransk, J., Thompson, M. E., Meng, G., & Galea, S. (2023) · The Annals of Applied Statistics · doi:10.1214/23-aoas1729
Aghabazaz, Z., Kazemi, I., & Nematollahi, A. R. (2023) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2023.115579
Westfall, H. A., & Lee, M. D. (2023) · Computational Brain & Behavior · doi:10.1007/s42113-023-00182-5
Tudor, J. (2023) · Cronfa (Swansea University) · doi:10.23889/suthesis.64642
Investigating the automaticity of links between body perception and trait concepts
Wildman, A., & Ramsey, R. (2023) · Visual Cognition · doi:10.1080/13506285.2023.2250505
“Also, by using weakly informative priors, we allow for the possibility of larger e ff ects, should they exist in the data (Gelman, 2006; Gelman et al., 2013; Gelman & Hill, 2006; Lemoine, 2019).”
MethodBayesian Inference and Simulation for Logarithm Transformed (LT) Method Using Stan
AbuJarad, M. H. A., AbuJarad, E. S. A., & Khan, A. A. (2022) · Journal of Modern Applied Statistical Methods · doi:10.56801/jmasm.v21.i2.2
Esin, A. V. C. I., & HAYAT, E. (2023) · doi:10.59287/nfnsm.837
Bayesian hierarchical modeling in interim futility analysis for two parallel clinical trials
Li, H., Roy, D., & Deng, Q. (2023) · Statistics in Medicine · doi:10.1002/sim.9971
Capture-Recapture: Bayesian Methods
Schofield, M. (2023) · Statistics for biology and health · doi:10.1007/978-3-031-39834-6_13
A flexible Bayesian hierarchical quantile spatial model for areal data
Fernandez, R. C., Gonçalves, K. C. M., & Pereira, J. B. D. M. (2023) · Statistical Modelling · doi:10.1177/1471082x231204930
“Following Gelman (2006), we set a folded-t distribution with 5 degrees of freedom for σ q , that is, σ q ∼ t (0 , ∞ ) (5 , 0 , 100), aiming to restrict the sampling of unusually large values.”
Art Knowledge Training Shapes Understanding, Inspires Creativity and Stimulates ThinkingPreprint
Bara, I., Ramsey, R., & Cross, E. S. (2024) · doi:10.31234/osf.io/sxqgc
Münch, F. F., & Koch, T. (2024) · doi:10.31234/osf.io/zpgtn
Beinhauer, L. J., Fuenderich, J., & Renkewitz, F. (2024) · doi:10.31234/osf.io/ud9rb
Zahrieh, D. M., Kandler, B. W., & Le‐Rademacher, J. G. (2024) · Clinical Trials · doi:10.1177/17407745231219680
Lesage, V. S., Wing, S., Zuur, A. F., et al. (2024) · Frontiers in Marine Science · doi:10.3389/fmars.2024.1289220
Hegde, S. T., Eisenberg, J. N. S., Beesley, L. J., & Mukherjee, B. (2024) · medRxiv · doi:10.1101/2024.03.13.24304191
Tractable Bayesian Inference For An Unidentified Simple Linear Regression Model
Jump, R. C. (2024) · The American Statistician · doi:10.1080/00031305.2024.2333864
“Using the same beta-distributed Q , it is also the case that, in which the generalized beta prime distribution, GenBetaPrime (α , η , κ , λ , μ) , has the probability density function, , which is a half-t distribution with ν degrees of freedom (Gelman 2006; Crooks 2019).”
The evolution of the mass - X-ray luminosity relation in the CODEX galaxy cluster sample
Pederneiras, I. (2024) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/d.14.2024.tde-26032024-191951
Estimation of world seroprevalence of SARS-CoV-2 antibodies
Lee, K.-M., Kim, S.-M., Jo, S., & Lee, J. (2024) · Journal of Applied Statistics · doi:10.1080/02664763.2024.2335569 · Cited by 1
“Gelman et al. (2006) suggested the flat prior for the standard deviation σ in hierarchical models, and they also showed that this prior gives the proper posterior distribution when flat priors are given for other parameters, μ0, τ , β (I) 1 and β (I) 2 for our model. For p + l and p − l , we construct prior distributi…”
Bargaining between the sexes: outside options and leisure time in hunter-gatherer householdsPreprint
Deb, A., Saunders, D., Major‐Smith, D., et al. (2024) · doi:10.31219/osf.io/vqzta
Keller, E., Lawson, A. B., Wagner, C. L., et al. (2024) · BMC Medical Research Methodology · doi:10.1186/s12874-024-02211-8
“In addition, since the marginal distribution for the variance of an inverse-Wishart is an inverse gamma distribution, datasets in which low variances are plausible can yield sensitive and biased inferences [23].”
BackgroundBayesian Multisource Hierarchical Models with Applications to the Monthly Retail Trade Survey
Kaputa, S. J., Morris, D. S., & Holan, S. H. (2024) · Journal of Survey Statistics and Methodology · doi:10.1093/jssam/smae019 · Cited by 1
Approximate inferences for Bayesian hierarchical generalised linear regression models
Berman, B., Johnson, W. O., & Shen, W. (2024) · Australian & New Zealand Journal of Statistics · doi:10.1111/anzs.12412
Batinović, L., Hammarsten, J., Nilsson, K. B., Topor, M. K., Holmer, E., Parlett-Pelleriti, C., & Danielsson, H. (2024) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/xu536
Costs of Multimorbidity and Patient Lived Experience (COMPLEX)
Tran, P. B. (2024) · doi:10.63028/10067/2061810151162165141 · Cited by 1
Chinembiri, T. S. (2023) · ResearchSpace (University of KwaZulu-Natal) · doi:10.29086/10413/23148
Yu, J., & Baroud, H. (2024) · Journal of Statistical Computation and Simulation · doi:10.1080/00949655.2024.2364843
“[28], when the group-level variance is close to zero, must be set to a reasonable value.”
(2024) · Elsevier eBooks · doi:10.1016/b978-0-443-13715-0.00020-0
Nam, O. V., Thạch, N. N., & Hải, N. M. (2024) · Studies in systems, decision and control · doi:10.1007/978-3-031-59110-5_14
Mapping socio-economic status using mixed data: a hierarchical Bayesian approach
Virgili-Gervais, G., Schmidt, A. M., Bixby, H., Cavanaugh, A., Owusu, G., Agyei‐Mensah, S., Robinson, B. E., & Baumgartner, J. C. (2024) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1093/jrsssa/qnae080
“Let σ = ( σ 1 , …, σ p ) ′ be the vector of the standard deviations of ϵ ijk , for k = 1, 2, …, p, we assign a half-Cauchy prior, that is, σ k ~ Half-Cauchy(0, 1) (Gelman, 2006) for continuous variables and assume it to be fixed at one for dichotomous variables (Quinn, 2004; Schliep & Hoeting, 2013).”
InfluentialNishio, M., & Ota, E. (2024) · PeerJ Computer Science · doi:10.7717/peerj-cs.2380 · Cited by 1
“The preference for the truncated normal distribution over the gamma distribution for the prior distribution of discrimination parameters is supported by previous studies ( Röver et al., 2021 ; Gelman, 2006 ), hence the truncated normal distribution was adopted in the present study.”
Dynamic Heterogeneity in Discrete Choice ExperimentsPreprint
Hagemann, N., Guhl, D., Kneib, T., Möllenhoff, K., & Steiner, W. J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4957076
Honey, I Shrunk the Irrelevant Effects! Simple and Fast Approximate Bayesian RegularizationPreprint
Karimova, D., van Erp, S., Leenders, R., & Mulder, J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.5004107 · Cited by 1
Unfolding the Network of Peer Grades: A Latent Variable Approach
Mignemi, G., Chen, Y., & Moustaki, I. (2025) · Psychometrika · doi:10.1017/psy.2025.10021 · Cited by 3
Can energy density influence treefrog distribution and species functional traits?
Lourenço‐de‐Moraes, R., Maciel, A. L. F., Ferreira, R. B., & Benedito, E. (2024) · Evolutionary Ecology · doi:10.1007/s10682-024-10319-y · Cited by 1
Balayla, J. (2024) · Springer eBooks · doi:10.1007/978-3-031-71452-8_3
Nolan, T. H., Richardson, S. T., & Ruffieux, H. (2024) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2024.108094 · Cited by 6
“This hierarchical construction based on auxiliary variables facilitates arbitrarily non-informative priors on standard deviation parameters (Gelman, 2006).”
Qian, Y., & Ma, B. (2024) · medRxiv · doi:10.1101/2024.12.19.24319363
Sugano, Y. (2024) · Consciousness and Cognition · doi:10.1016/j.concog.2024.103806
A Bayesian Binomial Regression Model for Ozone Levels in Northern Italy
Frigeri, M., Marchesin, L., Coppellotti, M., et al. (2025) · Springer eBooks · doi:10.1007/978-3-031-64447-4_7
Rahman, A., Jiang, D., & Lix, L. M. (2025) · British Journal of Mathematical and Statistical Psychology · doi:10.1111/bmsp.12381
Leontyeva, Y., Huang, Y., Cramb, S. M., Cameron, J. K., Baade, P. D., Mengersen, K. L., & Thompson, H. (2025) · Statistics in Medicine · doi:10.1002/sim.10287 · Cited by 1
“First, the standard deviations were given a Uniform prior distribution (hyperprior 1), as recommended by Gelman [37] to minimize the information it provides.”
Necessary and Sufficient Conditions for Posterior Propriety for Generalized Linear Mixed Models
Rao, Y., & Roy, V. (2025) · Sankhya A · doi:10.1007/s13171-025-00376-y
Semi-Parametric Item Response Theory With O’Sullivan Splines for Item Responses and Response Time
Liu, C. (2025) · Applied Psychological Measurement · doi:10.1177/01466216251316277 · Cited by 4
“To draw samples from a half-Cauchy distribution, the last line of equation (9) is translated into where the inverse gamma distribution is used, with A u = A ε = 25; thus, the prior information is vague (Gelman, 2006).”
Verde, P. E., & Rosner, G. L. (2025) · Biometrical Journal · doi:10.1002/bimj.70034 · Cited by 2
Using prior-data conflict to tune Bayesian regularized regression models
Biziaev, T., Kopciuk, K. A., & Chekouo, T. (2025) · Statistics and Computing · doi:10.1007/s11222-025-10582-1 · Cited by 2
Garraza, L. G., Cardona, C., Gichangi, P., Thiongo, M. N., Anglewicz, P. A., & Alkema, L. (2025) · Health Services and Outcomes Research Methodology · doi:10.1007/s10742-025-00342-z
“| ( ) For Bayesian estimation we need to advance priors for S , S and B , B , we use flat improper prior for the regression coefficients and weakly informative prior for the variance component (Gelman 2006).”
InfluentialBayesian modeling with locally adaptive prior parameters in small animal imaging
Zhang, M., Aykroyd, R. G., & Tsoumpas, C. (2025) · Frontiers in Nuclear Medicine · doi:10.3389/fnume.2025.1508816
“For the fi rst method, the uniform prior allocates the same probability to each value within its de fi ned range; it is subjective but hard to estimate values when the value is extremely small (23, 24).”
Hypothesis Testing For Dependent Variables With Unbalanced DataPreprint
Julián, J. M. G. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5172185
Computationally efficient spatio-temporal disease mapping for big data
Lee, D. (2025) · Spatial Statistics · doi:10.1016/j.spasta.2025.100901
Lin, Y.-S., Hsieh, N., Schlosser, P. M., Dzierlenga, M. W., & Ju, H. (2025) · Toxicological Sciences · doi:10.1093/toxsci/kfaf070
Bara, I., Schöni, L., & Cross, E. S. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5262324
Generalization of Rejection and Acceptance in Social NetworksPreprint
Zhang, Y., & Hackel, L. M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5263641
WITHDRAWNPreprint
Leizerman, S. L. (2025) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/aekqs_v1
Tipping point analysis in network meta-analysis
Wang, Z., Murray, T. A., Han, W., Lin, L., Siegel, L. K., & Chu, H. (2025) · Research Synthesis Methods · doi:10.1017/rsm.2025.24
Daniells, L., Mozgunov, P., Barnett, H., Bedding, A. W., & Jaki, T. (2024) · Biostatistics · doi:10.1093/biostatistics/kxaf016 · Cited by 1
“Gelman (2006) argued that the original suggestion of an Inverse-Gamma prior by Berry et al. (2013) had poor behavior when σ 2 is too close to 0, thus suggested a Half-Cauchy prior instead.”
BiGER: Bayesian rank aggregation in genomics with extended ranking schemes
Wang, K., Yang, Y., Xia, Y., Xiao, G., Lim, J., & Wang, X. (2026) · Nature Communications · doi:10.1038/s41467-026-72758-1 · Cited by 1
“We performed a parameter tuning exercise using a range of IG parameters as previously suggested in the literature 6,37,38 and found that there was no signi fi cant difference in coverage and Spearman ’ s correlation (Supplementary Fig.”
Efficient prior specification in procedural 3D modelling
Kleitsiotis, I., Tsirogiannis, G. A., & LIKOTHANASSIS, S. D. (2025) · Simulation Modelling Practice and Theory · doi:10.1016/j.simpat.2025.103165
VEHICLE AND SECTION-DEPENDENT SPEED-SPACING RELATION ESTIMATION USING HIERARCHICAL BAYESIAN MODEL
HAYASHI, K., & Seo, T. (2024) · Japanese Journal of JSCE · doi:10.2208/jscejj.24-20029
Mhorseshoe package in R: Approximate algorithm for the horseshoe prior in Bayesian linear model
Kang, M., & Lee, K. (2025) · SoftwareX · doi:10.1016/j.softx.2025.102236
Bayesian Hierarchical Mixture Models for Adaptive Basket Trial Design
Chen, C., Chang, L.-C., & Hsiao, C. (2025) · Statistics in Biopharmaceutical Research · doi:10.1080/19466315.2025.2529427
“Selection of the hyperprior for σ 2 is crucial in BHMs because the hyperprior usually determines the strength of information borrowing (Gelman 2006).”
Bayesian time-varying meta-analysis via hierarchical mean-variance random-effects models
Kubota, K., Sugasawa, S., Ochiai, K., et al. (2025) · Japanese Journal of Statistics and Data Science · doi:10.1007/s42081-025-00311-3
“As the default choice, we adopt relatively standard prior distributions as follows. where HC( η ) denotes the half-Cauchy prior (e.g. Gelman, 2006) with scale parameter η .”
Consistent group selection using global–local shrinkage priors in sparse normal linear regression
Paul, S., Ghosh, P., & Chakrabarti, A. (2025) · Annals of the Institute of Statistical Mathematics · doi:10.1007/s10463-025-00950-z · Cited by 1
“As a result, from (5), it is clear that the variance of ̂𝜷 g remains on the same scale as the prior distribution of g in line with the suggestion of Gelman (2006).”
Wang, S., Roychoudhury, S., & Kidwell, K. (2025) · Biometrics · doi:10.1093/biomtc/ujaf103
Metadata
- Authors
- Gelman, A.
- Affiliations
- Columbia University (US)
- Corresponding author
- Gelman, Andrew
- ORCID
- Gelman, Andrew
- Venue
- Bayesian Analysis
- Publisher
- Institute of Mathematical Statistics
- Published
- 2006-09-01
- Volume
- 1
- Issue
- 3
- Pages
- 515-534
- ISSN
- 1931-6690 (ISSN-L)
- 1936-0975
- OpenAlex
- W1536497620
- Semantic Scholar
- 63d40805d4635b70f6b5aa07fc2c9821480cfc1e
- Cited by
- 4,050
- References
- 34
- Citations by year
- 2026: 194 · 2025: 218 · 2024: 237 · 2023: 237 · 2022: 282 · 2021: 312 · 2020: 320 · 2019: 281 · 2018: 266 · 2017: 263 · 2016: 235 · 2015: 217 · 2014: 176 · 2013: 142 · 2012: 184
- FWCI
- 78.70
- Type
- journal article
- Language
- en
- Open access
- diamond
- doi.org
- Licence
- cc-by
- Locations
- Bayesian Analysis (publishedVersion, cc-by)
- Project Euclid (Cornell University) (submittedVersion)
- Funders
- National Science Foundation
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Distribution Estimation and Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Bayesian Methods and Mixture Models
- Keywords
- prior distributions, hierarchical models, scale parameter, non-informative priors, weakly informative priors, uniform prior, analysis of variance
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-02, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-02 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.