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 2,901–3,000 of 4,089
Olarinre, M., Siegle, J. H., & Kass, R. E. (2026) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2025.1715136 · Cited by 2
“We place a half-Cauchy distribution on the standard deviation, with scale parameter 1 (Polson and Scott, 2012; Gelman, 2006).”
Swinging, Fast and Slow: Interpreting Variation in Baseball Swing Tracking Metrics
Powers, S., & Yurko, R. (2026) · The American Statistician · doi:10.1080/00031305.2026.2633338 · Cited by 2
Scheidt, D. M., Provancha, J. A., Stolen, E. D., Breininger, D. R., Cancro, R. A., Lowers, R. H., Reyier, E. A., & Ahr, B. J. (2026) · Frontiers in Marine Science · doi:10.3389/fmars.2026.1725777 · Cited by 2
Hu, C., Zhang, J., Ren, W., & Liang, H. (2026) · Frontiers of Architectural Research · doi:10.1016/j.foar.2026.01.014 · Cited by 3
An Overview of the Application of Modern Statistical Techniques in Semiconductor Manufacturing
Chen, H., & Chen, C. H. (2026) · Applied System Innovation · doi:10.3390/asi9040083 · Cited by 4
Hierarchical Models and More on Convergence Assessment
Cowles, M. K. (2012) · Springer texts in statistics · doi:10.1007/978-1-4614-5696-4_9 · Cited by 2
Agapova, M., Devine, E. B., Nguyen, H. P., Wolf, F. M., & Inoue, L. Y. T. (2014) · Journal of Comparative Effectiveness Research · doi:10.2217/cer.14.29 · Cited by 2
A note on the existence of the posteriors for one-way random effect probit models
Lin, X., & Sun, D. (2009) · Statistics & Probability Letters · doi:10.1016/j.spl.2009.09.012 · Cited by 1
Dynamic Estimation of Latent Opinion from Sparse Survey Data Using a Group-Level IRT Model
Caughey, D., & Warshaw, C. (2013) · Cited by 2
Bayesian truncated beta nonlinear mixed-effects models
Paraíba, C. C. M., Bochkina, N., & Diniz, C. A. R. (2017) · Journal of Applied Statistics · doi:10.1080/02664763.2016.1276891 · Cited by 2
Li, X., Liu, X., & Chen, Y. (2018) · Asian-Australasian Journal of Animal Sciences · doi:10.5713/ajas.18.0102 · Cited by 2
JMASM 51: Bayesian Reliability Analysis of Binomial Model – Application to Success/Failure Data
Akhtar, M. T., & Khan, A. A. (2019) · Journal of Modern Applied Statistical Methods · doi:10.22237/jmasm/1553803862 · Cited by 2
“The half-Cauchy distribution with scale α = 25 is a recommended, default, weakly informative prior distribution for a scale parameter (Gelman, 2006).”
BackgroundBayesian inference of material properties in disordered media using sound characteristics
Kumar, A., Shrivastava, R., & Singh, K. H. (2020) · Europhysics Letters (EPL) · doi:10.1209/0295-5075/129/24001 · Cited by 2
Bayesianische Inferenz als Alternative zur klassischen Statistik
Gill, J., & Traunmüller, R. (2020) · Springer eBooks · doi:10.1007/978-3-658-16936-7_33 · Cited by 2
Draney, M. L., Doll, J. C., Doerr, L. R., Houghton, C. J., & Forsythe, P. S. (2020) · Environmental Entomology · doi:10.1093/ee/nvaa051 · Cited by 3
Maximum Probability Theorem: A Framework for Probabilistic Machine Learning
Marvasti, A. E., Marvasti, E. E., Bagci, U., et al. (2021) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2021.3086046 · Cited by 2
“Laplace’s Principle of Indifference [15], Conjugate Priors [9, 10], Principle of Maximum Entropy [14], Jeffreys priors [16] and Reference Priors [3].”
MethodThe role of credit and housing shocks in emerging economies
Gomes, F. A. R., & Soave, G. P. (2020) · Applied Economics Letters · doi:10.1080/13504851.2020.1830935 · Cited by 2
“As suggested by Gelman (2006), we set the hyperparameters s ¼ 0 and v ¼ 1.”
MethodNyamaruze, P., Cowden, R. G., Padgett, R. N., & Govender, K. (2021) · BMJ Open · doi:10.1136/bmjopen-2020-045700 · Cited by 2
Bayesian Hierarchical Factor Analysis for Efficient Estimation Across Race/Ethnicity
Hu, J., Clark, L. A., Shi, P., Staggs, V. S., Daley, C. M., & Gajewski, B. J. (2021) · Revista Colombiana de Estadística · doi:10.15446/rce.v44n2.87690 · Cited by 2
Tail Bounds forℓ1Norm of Gaussian Random Matrices with Applications
Gao, X., Zhang, M., & Luo, J. (2022) · Journal of Mathematics · doi:10.1155/2022/1456713 · Cited by 2
Speakers balance their use of cues to grammatical functions in informative discourse contexts
Hörberg, T., & Sjons, J. (2022) · Language Cognition and Neuroscience · doi:10.1080/23273798.2022.2102667 · Cited by 4
“In line with recommendations (e.g. Gelman, 2006; Gelman et al., 2008; Stan Development Team, 2017), standard, weakly regularising priors were used in the models.”
ResultBruch, C., & Felderer, B. (2022) · Austrian Journal of Statistics · doi:10.17713/ajs.v51i4.1361 · Cited by 3
Weber, F., Ickstadt, K., & Glass, A. S. (2022) · The R Journal · doi:10.32614/rj-2022-027 · Cited by 3
Bayesian historical borrowing with longitudinal large-scale assessments
Kaplan, D. E., Chen, J., Lyu, W., & Yavuz, S. (2023) · Large-scale Assessments in Education · doi:10.1186/s40536-022-00140-w · Cited by 7
Chen, X., Zhang, J., Jiang, L., et al. (2023) · BMC Medical Research Methodology · doi:10.1186/s12874-023-01877-w · Cited by 2
“For example, when the number of subgroups is small, it may be inappropriate to directly use the inverse gamma distribution with small shape and scale parameters as a prior for the variances across subgroups [31].”
MethodPaulon, G., Müller, P., & Sarkar, A. (2023) · Bayesian Analysis · doi:10.1214/23-ba1383 · Cited by 7
A Gibbs sampler for multi-species occupancy models
Clark, A. E., & Altwegg, R. (2023) · Environmental and Ecological Statistics · doi:10.1007/s10651-023-00558-7 · Cited by 2
“Finally, and are modelled using a half-Cauchy distribution (Gelman 2006) and can be specified hierarchically as where A and B are known hyper-parameters.”
MethodDirichlet process mixture models using matrix‐generalized half‐t distribution
Lee, S., & Kim, C. (2023) · Stat · doi:10.1002/sta4.599 · Cited by 2
“As various theories and methodologies for non-informative priors have developed over time (Datta & Ghosh, 1995, 1996; Daniels, 1999; Gelman, 2006), it is important to establish non-informative priors when conducting research with limited prior knowledge.”
BackgroundMethodInfluentialGlobal-Local Priors for Spatial Small Area Estimation
Tang, X., & Ghosh, M. (2023) · Calcutta Statistical Association Bulletin · doi:10.1177/00080683231186378 · Cited by 2
“Recently, the half-Cauchy prior has been recommended as a default prior for scale parameters as opposed to the inverse gamma priors IG( a 0 , b 0 ) (Gelman [9] ; Polson and Scott [18] ) as the performance with the latter prior may be sensitive to the choice of a 0 and b 0 .”
MethodWarren, J. L., Sundaram, M. E., Pitzer, V. E., et al. (2023) · Epidemiology · doi:10.1097/ede.0000000000001690 · Cited by 2
Jokhio, S., Olleja, P., Bärgman, J., Yan, F., & Baumann, M. R. K. (2023) · IET Intelligent Transport Systems · doi:10.1049/itr2.12457 · Cited by 3
Network meta analysis to predict the efficacy of an approved treatment in a new indication
Proper, J. L., Chu, H., Prajapati, P., Sonksen, M. D., & Murray, T. A. (2023) · Research Synthesis Methods · doi:10.1002/jrsm.1683 · Cited by 3
“35,36 The U(0,5) prior reflects the belief that the true value for σ is equally likely to take on any value between 0 and 5 and places 95 % of its probability mass between 0.12 and 4.88.”
BackgroundInfluentialRejoinder to “Specifying Prior Distribution in Reliability Applications”
Tian, Q., Lewis‐Beck, C., Niemi, J., & Meeker, W. Q. (2023) · Applied Stochastic Models in Business and Industry · doi:10.1002/asmb.2832 · Cited by 2
SHARING INFORMATION BETWEEN MACHINE TOOLS TO IMPROVE SURFACE FINISH FORECASTING
CLARKSON, D. R., Bull, L. A., Dardeno, T. A., et al. (2023) · International Workshop on Structural Health Monitoring (IWSHM) · doi:10.12783/shm2023/37063 · Cited by 2
DOD-Combo: Bayesian dose finding design in combination trials with meta-analytic-predictive prior
Chen, K., Zhao, Y., Liu, M., Lin, J., & Liu, R. (2024) · Journal of Biopharmaceutical Statistics · doi:10.1080/10543406.2024.2325142 · Cited by 3
“We assume that σ 2 follows a half-t prior distribution with one degree of freedom and scale parameter 25 (Gelman 2006) and that θ follows a prior distribution N J 0 ; 100 I J ð Þ .”
Inferring stochastic group interactions within structured populations via coupled autoregression
McGrane-Corrigan, B., Mason, O., & Moral, R. D. A. (2024) · Journal of Theoretical Biology · doi:10.1016/j.jtbi.2024.111793 · Cited by 2
“According to Gelman (2006) if one was to set a variance prior such as Inv-Γ( ϵ, ϵ ) , for ϵ small, then there still does not exist a proper (integrates to unity) limiting posterior distribution.”
MethodSmall Sample Methods in Multilevel Analysis
Yamamoto, Y., & Miyazaki, Y. (2024) · The Journal of Experimental Education · doi:10.1080/00220973.2024.2329084 · Cited by 2
A joint penalized spline smoothing model for the number of positive and negative COVID-19 tests
Witte, D. D., Alonso, A., Neyens, T., Verbeke, G. N., & Molenberghs, G. (2024) · PLoS ONE · doi:10.1371/journal.pone.0303254 · Cited by 4
Familial aggregation of seizure outcomes in four familial epilepsy cohorts
Ellis, C. A., Tu, D., Oliver, K. L., Mefford, H. C., Hauser, W. A., Buchhalter, J., Epstein, M. P., Cao, Q., Berkovic, S. F., & Ottman, R. (2024) · Epilepsia · doi:10.1111/epi.18004 · Cited by 2
“42,43 In sensitivity analyses, we used two alternative priors: a half-normal distribution with mean 0 and standard deviation (SD) 1, which places less probability mass on higher values; and a half-Cauchy distribution, which places more probability mass on higher values.”
Byrne, C., Coetzer, R., & Ramsey, R. (2024) · Archives of Clinical Neuropsychology · doi:10.1093/arclin/acae032 · Cited by 2
“We set priors using a weakly informative approach (Gelman, 2006).”
Vieira, F., Leenders, R., & Mulder, J. (2024) · Journal of Computational Social Science · doi:10.1007/s42001-024-00290-7 · Cited by 2
“In this case, the posterior would be completely determined by the likelihood.”
BackgroundExamining time-varying dynamics of co-occurring depressed mood and anxietyPreprint
Piccirillo, M. L., Frumkin, M. R., Spink, K. M., Tonge, N., & Foster, K. T. (2021) · PsyArXiv (OSF Preprints) · doi:10.1016/j.jad.2024.06.064 · Cited by 4
Kim, J., Kang, S.-H., Jin, I. H., & Park, J. (2024) · Communications in Statistics - Simulation and Computation · doi:10.1080/03610918.2024.2371003 · Cited by 2
“…paper was to investigate the change in results according to the types of prior distributions of s 2 : Referring to several Bayesian studies (e.g. Gelman 2006; Gelman and Hill 2006; Gustafson, Hossain, and Macnab 2006; Polson and Scott 2012; Hsu, Zalkikar, and Tiwari 2019), we considered two…”
Alharthi, A. S., & Almulhim, F. A. (2024) · Alexandria Engineering Journal · doi:10.1016/j.aej.2024.05.094 · Cited by 2
Design and Structure Dependent Priors for Scale Parameters in Latent Gaussian Models
Gardini, A., Greco, F. P., & Trivisano, C. (2024) · Bayesian Analysis · doi:10.1214/24-ba1454 · Cited by 2
“As pointed out by Gelman (2006) and Polson & Scott (2012), selecting priors for scale parameters of hierarchical models requires particular attention since such parameters are not directly informed by data.”
BackgroundMethodInfluentialSchibber, C. F. (2024) · Legislative Studies Quarterly · doi:10.1111/lsq.12469 · Cited by 3
“…Latin American democracies: Argentina (1998–2009), Bolivia (1997–2010), Chile (1997–2010), Colombia (1998–2010), Costa Rica (1998–2010), El Salvador (1997–2006), Guatemala (1995–2008), Honduras (1997–2010), Mexico (1997–2006), Nicaragua (1996– 2011), Paraguay (1998–2008), and Uruguay (2000– 2010).”
Multidisciplinary considerations for implementing Bayesian borrowing in basket trials
Broglio, K. R., Blau, J. E., Pilling, E. A., et al. (2024) · Drug Discovery Today · doi:10.1016/j.drudis.2024.104127 · Cited by 2
Hill, M. P., Caley, P. A., Stuart, I., Duncan, R. P., & Forsyth, D. M. (2024) · Biological Invasions · doi:10.1007/s10530-024-03409-z · Cited by 5
“For priors on variances, following Gelman (2006) we used a uniform prior on the scale of the standard deviation.”
Bayesian covariance structure modeling of interval-censored multi-way nested survival data
Baas, S., Fox, J., & Boucherie, R. J. (2024) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2024.105359 · Cited by 2
Age of Peak Performance Among Soccer Players in Sweden
Säfvenberg, R., Nordgaard, A., Eriksson, O. L., et al. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-69073-0_24 · Cited by 2
“Similarly, the priors for ( τ µ , τ σ ) were also set as weakly informative based on [11].”
BackgroundIdeational diffusion and the great witch hunt in Central Europe
Doten-Snitker, K., Pfaff, S. J., & Hsiao, Y. (2024) · Theory and Society · doi:10.1007/s11186-024-09576-1 · Cited by 3
“We use city-specific random intercepts with the narrower Student’s half-t distribution (df = 4, scale = 1) for the intercept priors (Gelman, 2006) because we do not expect much city-spe-cific variation.”
Can the hot hand phenomenon be modelled? A Bayesian hidden Markov approach
Calvo, G., Armero, C., & Spezia, L. (2024) · Computational Statistics · doi:10.1007/s00180-024-01560-8 · Cited by 7
“However, Gelman (2006), and Martínez-Beneito and Botella-Rocamora (2019) argue that the gamma distribution exhibits undesirable behaviour (relative to its percentiles) or converges to an improper distribution, resulting in an improper posterior distribution.”
Yao, M., Mei, F., Zou, K., Li, L., & Sun, X. (2024) · Pharmaceutical Statistics · doi:10.1002/pst.2448 · Cited by 3
Poli, G., Argiento, R., Amedei, A., et al. (2024) · Biometrical Journal · doi:10.1002/bimj.70000 · Cited by 2
“For the standard deviation of the random effects, 𝜆 and 𝜏 𝑗 , in the absence of strong a priori information, following Gelman (2006), we assume a uniform prior.”
Modeling Survey Time Series Data with Flow-Observed CARMA Processes
Joyce, P. M., & McElroy, T. (2024) · Journal of Official Statistics · doi:10.1177/0282423x241286236 · Cited by 2
Robust modeling for continuous bounded spatial data
Ahmadi, A., Khaledi, . J., Sohrabi, H., et al. (2025) · Stochastic Environmental Research and Risk Assessment · doi:10.1007/s00477-025-02919-z · Cited by 2
A Splitting Criterion for CART Models Based on Bayesian Optimisation
de Paz, E. G. G., Vaquera‐Huerta, H., Velasco, F. J. A., et al. (2024) · Springer proceedings in mathematics & statistics · doi:10.1007/978-3-031-78401-9_2 · Cited by 2
A new conservative and robust Bayesian approach for the event date model in chronology building
Noukpoape, K. M., Lanos, P., & Dufresne, P. (2025) · Archaeometry · doi:10.1111/arcm.13063 · Cited by 4
Treatment-Control Comparisons in Platform Trials Including Non-Concurrent Controls
Roig, M. B., Krotka, P., Hees, K., et al. (2025) · Statistics in Biopharmaceutical Research · doi:10.1080/19466315.2025.2496316 · Cited by 2
“To calibrate such a weakly informative prior, it can be helpful to think about plausible changes in E ( Y ) between time buckets c and c − 1.”
Generalized Decomposition Priors on R2
Aguilar, J. E., & Bürkner, P. (2025) · Bayesian Analysis · doi:10.1214/25-ba1524 · Cited by 6
Boamah, N. A., Ofori-Yeboah, F., & Appiagyei, K. (2025) · Asian Journal of Economics and Banking · doi:10.1108/ajeb-03-2024-0032 · Cited by 2
Poonai, N., Arthur‐Hayward, V., Ali, S., Sabhaney, V. J., Doan, Q. H., Trottier, E. D., Gravel, J., Tran, N. A., Bhatt, M., Eltorki, M., et al. (2025) · PLoS ONE · doi:10.1371/journal.pone.0324515 · Cited by 2
“We will use predefined minimally informative priors for other analyses [96,97] along with descriptive statistics.”
Woldegerima, W. A., & Ugwu, C. L. J. (2025) · Infectious Disease Modelling · doi:10.1016/j.idm.2025.07.011 · Cited by 5
“This provides mild regularization while retaining flexibility in estimating the log-relative risk scale (Gelman, 2006).”
Iacovone, L., McKenzie, D. J., & Meager, R. (2025) · Econometrica · doi:10.3982/ecta21567 · Cited by 4
Bayesian multivariate track geometry degradation modeling and its use in condition-based inspection
Ba, H. T., Rebello, S., Cholette, M. E., Reddy, V., & Borghesani, P. (2025) · Railway Engineering Science · doi:10.1007/s40534-025-00394-4 · Cited by 6
Borden, J. B., Gibbs, J. P., Vanek, J. P., & Cosentino, B. J. (2025) · Urban Ecosystems · doi:10.1007/s11252-025-01841-0 · Cited by 2
Synthesis and perceptual scaling of high-resolution naturalistic images using Stable Diffusion
Pettini, L., Bogler, C., Doeller, C. F., & Haynes, J. (2025) · Behavior Research Methods · doi:10.3758/s13428-025-02889-8 · Cited by 4
“016).”
Application of Interweaving in DLMs to an Exchange and Specialization Experiment
Simpson, M. J. (2015) · Springer proceedings in mathematics & statistics · doi:10.1007/978-3-319-16238-6_7 · Cited by 4
Ball, S. J., & Pereira, G. F. (2013) · Applied Geography · doi:10.1016/j.apgeog.2013.03.018 · Cited by 2
A global-local approach for detecting hotspots in multiple-response regression
Ruffieux, H., Davison, A. C., Hager, J., Inshaw, J. R. J., Fairfax, B. P., Richardson, S. T., & Bottolo, L. (2020) · The Annals of Applied Statistics · doi:10.1214/20-aoas1332 · Cited by 16
“Gelman (2006) discusses the relevance of several noninformative and weakly informative priors on random effect variances.”
BackgroundMiten ja milloin yksilön kieli muuttuu?
Lappalainen, H., Mustanoja, L., & O’Dell, M. L. (2019) · Virittäjä · doi:10.23982/vir.67808 · Cited by 2
Willink, B., Blow, R., Sparrow, D. J., et al. (2021) · Ecological Entomology · doi:10.1111/een.13006 · Cited by 2
“These priors come from a non-central scaled F-distribution, and we used a scale of 25 following Gelman (2006).”
MethodJaya, I. G. N. M., Andriyana, Y., & Tantular, B. (2022) · Geospatial health · doi:10.4081/gh.2022.1070 · Cited by 2
Ueda, M. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-05064-0_28 · Cited by 2
Erciulescu, A. L., Li, J., Krenzke, T., et al. (2022) · Statistical Methods & Applications · doi:10.1007/s10260-022-00678-7 · Cited by 2
“See Krenzke et al. (2020), Browne and Draper (2006), Gelman (2006), and Polson and Scott (2012), for some recent discussions about the choice of prior distribution for the scale parameter in a univariate hierarchical Bayes model.”
van den Bergh, D., Vandermeulen, N., Lesterhuis, M., Mæyer, S. D., Steendam, E. V., Rijlaarsdam, G., & van den Bergh, H. (2023) · Journal of Writing Research · doi:10.17239/jowr-2023.14.03.05 · Cited by 2
Introduction and Motivating Examples
Wakefield, J. (2012) · Springer series in statistics · doi:10.1007/978-1-4419-0925-1_1 · Cited by 2
Repeated measures and longitudinal data analysis using higher‐order entropies
Al‐Rawwash, M., & Al‐Nasser, A. D. (2012) · Statistica Neerlandica · doi:10.1111/j.1467-9574.2012.00534.x · Cited by 3
“Several studies point out that parametric models may get complicated to a point which an alternative modeling strategy becomes imminent (Lin and Carroll, 2001; Al-Rawwash, 2005; Gelman, 2006; Qin and Zhu, 2008; Buzkova and Lamley, 2008).”
BackgroundMølgaard, B., Vanhatalo, J. P., Aalto, P. P., et al. (2016) · Atmospheric measurement techniques · doi:10.5194/amt-9-741-2016 · Cited by 2
Statistical Modelling of Neighbor Treatment Effects in Aquaculture Clinical Trials
Masaoud, E., Stryhn, H., Whyte, S. K., et al. (2010) · Journal of Agricultural Biological and Environmental Statistics · doi:10.1007/s13253-010-0043-5 · Cited by 2
“In our view, the close link to the Wishart distribution (used for the MMD model) made the inverse variance gamma priors the most natural choice here, but the suggested uniform prior for random effect standard deviations ( Gelman 2006 ) was also explored in a sensitivity analysis.”
MethodWraith, D., Mengersen, K. L., Alston‐Knox, C., et al. (2014) · The Annals of Applied Statistics · doi:10.1214/13-aoas678 · Cited by 2
Bayesian semiparametric models for nonignorable missing mechanisms in generalized linear models
Kalaylıoğlu, Z., & Ozturk, O. (2013) · Journal of Applied Statistics · doi:10.1080/02664763.2013.794329 · Cited by 2
Lin, C. W., Warram, J. H., & Veng‐Pedersen, P. (2011) · Journal of Pharmacokinetics and Pharmacodynamics · doi:10.1007/s10928-011-9208-2 · Cited by 2
Bayesian assessment of times to diagnosis in breast cancer screening
Armero, C., López‐Quílez, A., & López-Sánchez, R. (2008) · Journal of Applied Statistics · doi:10.1080/02664760802191397 · Cited by 2
Thomson, J. R., Fonseca, D. M., Finelli, C. M., et al. (2015) · Freshwater Biology · doi:10.1111/fwb.12688 · Cited by 2
Comparison of inference methods of genetic parameters with an application to body weight in broilers
Maniatis, G., Demiris, N., Kranis, A., et al. (2015) · Archives animal breeding/Archiv für Tierzucht · doi:10.5194/aab-58-277-2015 · Cited by 2
Generalizing boundaries for triangular designs, and efficacy estimation at extended follow-ups
Allison, A., Edwards, T., Omollo, R., et al. (2015) · Trials · doi:10.1186/s13063-015-1018-1 · Cited by 2
“However, we do not choose a non-informative prior for τ , as this leads to problematic inferences when the number of groups is small [16].”
BackgroundPRODUCTION UNCERTAINTIES MODELLING BY BAYESIAN INFERENCE USING GIBBS SAMPLING
Azizi, A., Ali, A., Ping, L. W., & Mohammadzadeh, M. (2015) · The South African Journal of Industrial Engineering · doi:10.7166/26-3-572 · Cited by 2
“Another non-informative prior distribution sometimes proposed in the Bayesian literature is uniform, which is not recommended by Gelman [34] because of its miscalibration.”
BackgroundKalaylıoğlu, Z., & Demirhan, H. (2015) · Statistical Methods in Medical Research · doi:10.1177/0962280215615003 · Cited by 3
The Association between Urban Form and Ischemic Heart Disease: Evidence from Brisbane, Australia
Gudes, O., Ball, S. J., Dur, F., Burke, M., & Varhol, R. (2015) · eSpace (Curtin University) · doi:10.5897/jphe.9000125 · Cited by 2
Modelling multivariate, overdispersed binomial data with additive and multiplicative random effects
Fava, E. D., Shkedy, Z., Aregay, M., & Molenberghs, G. (2014) · Statistical Modelling · doi:10.1177/1471082x13503450 · Cited by 4
Polansky, A. M., & Maple, A. (2016) · Quality Technology & Quantitative Management · doi:10.1080/19369816.2016.1169687 · Cited by 2
“See Gelman (2006) for further information on the advantages of this type of hierarchical model.”
BackgroundCombining patient preferences with expected treatment outcomes to inform decision-making
Schuler, M. S., & Hatfield, L. A. (2016) · Health Services and Outcomes Research Methodology · doi:10.1007/s10742-016-0166-4 · Cited by 2
“…Grade 3–5 4 0 No/late recurrence* Recur C5 years after treatment 4 1 Recurrence Recur\5 years after treatment 5 0 Late death* Die C5 years after treatment 5 1 Death Die\5 years after treatment * Reference level of each outcome prior that is minimally informative and well-behaved (Gelman 2006).”
BackgroundDagne, G. A., & Ibrahimou, B. (2016) · Journal of Biopharmaceutical Statistics · doi:10.1080/10543406.2016.1269782 · Cited by 2
“…of prior distributions, we conducted a sensitivity analysis using a uniform prior distribution (unif ð0; 100Þ) for the standard deviation parameter σ (Gelman, 2006) instead of inverse gamma prior distribution (IGð0:01; 0:01Þ) for variance σ2, a gamma distribution (gammað3:5; 3:5ÞIð2; Þ) for the…”
MethodA Non-Stationary Spatial Model for Temperature Interpolation Applied to the State of Rio De Janeiro
Cunha, M., Gamerman, D., Fuentes, M., et al. (2017) · Journal of the Royal Statistical Society Series C (Applied Statistics) · doi:10.1111/rssc.12207 · Cited by 2
“A few alternative distributions were tested and good results were generally obtained with the half-Cauchy prior distributions that were suggested by Gelman (2006).”
MethodWu, Y., Gaskins, J., Kong, M., et al. (2017) · Biometrics · doi:10.1111/biom.12742 · Cited by 2
“Further, the heavy tail of the half-Cauchy prior protects from over-shrinkage by allowing large values when the effect is highly influential (Gelman, 2006; Carvalho, Polson, and Scott, 2010).”
BackgroundSahoo, M. M., & Patra, K. C. (2017) · Human and Ecological Risk Assessment An International Journal · doi:10.1080/10807039.2017.1394174 · Cited by 2
“The asymmetrical patterns of these dynamic traces justify the unidentified stochastic nodes of posterior distribution (Gelman 2006).”
BackgroundBayesian inference on contingency tables with uncertainty about independence for small areas
Kwak, S. G., Nandram, B., & Kim, D. H. (2017) · Journal of Applied Statistics · doi:10.1080/02664763.2017.1413074 · Cited by 2
Sparse Portfolio Selection via Bayesian Multiple Testing
Das, S., & Sen, R. (2020) · Sankhya B · doi:10.1007/s13571-020-00240-z · Cited by 2
Phillip, A., Chan, J., & Peiris, S. (2018) · Studies in Nonlinear Dynamics and Econometrics · doi:10.1515/snde-2015-0110 · Cited by 2
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.