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 2,501–2,600 of 4,089
Space–time regression modeling of tree growth using the skew‐t distribution
Nathoo, F. S. (2010) · Environmetrics · doi:10.1002/env.1057 · Cited by 5
Rasch Model and Its Extensions for Analysis of Aphasic Deficits in Syntactic Comprehension
Gutman, R., DeDe, G., Caplan, D. N., et al. (2011) · Journal of the American Statistical Association · doi:10.1198/jasa.2011.ap10017 · Cited by 5
“We also examined two other diffused priors (U[0,100] and half-Cauchy) as suggested by Gelman (2006) and found that the different prior distributions had little effect on the outcome.”
MethodMeta-analyses using information reweighting: An application to online advertising
Wang, P., Bradlow, E. T., & George, E. I. (2014) · Quantitative Marketing and Economics · doi:10.1007/s11129-014-9145-7 · Cited by 6
“Indeed, Bayesian modeling commonly utilizes non-informative or weakly informative priors (Gelman 2006; Gelman et al. 2008) as if external information was not available.”
MethodLikelihood inference for small area estimation using data cloning
Torabi, M., Lele, S. R., & Prasad, N. (2015) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2015.03.013 · Cited by 5
“To illustrate that the results of the HB method are not invariant to priors, we used uniform distribution U(0, 1000), instead of the gamma distribution as a prior for the standard error (Gelman, 2006).”
MethodKasakoff, A. B., Lawson, A. B., & Dressler, E. V. M. (2014) · Demographic Research · doi:10.4054/demres.2014.30.36 · Cited by 5
Sudderth, E. A., Clair, S. B. S., Placella, S., Swarbreck, S. M., Castanha, C., Herman, D. J., Fischer, M. L., Kleber, M., Sudderth, E. B., Torn, M. S., et al. (2012) · Ecosphere · doi:10.1890/es12-00004.1 · Cited by 5
González, L. (2016) · Studies in Comparative International Development · doi:10.1007/s12116-015-9205-3 · Cited by 5
“…2 (0.430 in model 1 and 0.124 in model 2) in these (as well as in the other) models not only justify the need of running a multilevel regression (as Gelman 2006: 524 suggests) but also can be interpreted as evidence that we still need better theories and predictors in the regressions due to…”
BackgroundInfluentialThe rest-frame Golenetskii correlation via a hierarchical Bayesian analysis
Burgess, J. M. (2017) · Monthly Notices of the Royal Astronomical Society · doi:10.1093/mnras/stx1159 · Cited by 5
“HalfCauchy distributions were chosen as non-informative proper priors for the variances (Gelman 2006).”
MethodSmith, A. N. H., & Anderson, M. J. (2016) · Ecology and Evolution · doi:10.1002/ece3.2406 · Cited by 5
“Following Gelman (2006), scaled, noncentral F- distributions were used as priors for the variance components for the random effects, with the scale parameter set to 1 for species richness and to 0.3 for the total number of fish.”
MethodInfluentialBayesian multi-antenna sensing in cognitive radio networks using Fractional Bayes Factor
Al-Ali, M. H., & Ho, K. C. (2017) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2017.7952822 · Cited by 5
“For the hypothesis that PU is not active, we shall use the conjugate prior for the unknown noise parameters, which unfortunately becomes improper as the hyperparameters go to zero [26].”
BackgroundMethodMielke, F., Schunke, V., Wölfer, J., & Nyakatura, J. A. (2018) · Zoology · doi:10.1016/j.zool.2018.05.009 · Cited by 5
A Bayesian two-stage spatially dependent variable selection model for space–time health data
Choi, J., & Lawson, A. B. (2018) · Statistical Methods in Medical Research · doi:10.1177/0962280218767980 · Cited by 5
Choi, J., & Levy, R. (2017) · Biometrics & Biostatistics International Journal · doi:10.15406/bbij.2017.06.00182 · Cited by 6
An objective prior for hyperparameters in normal hierarchical models
Berger, J. O., Sun, D., & Song, C. (2020) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2020.104606 · Cited by 5
Liu, J., Wick, J., Martin, R., Meinzer, C., Roy, D., & Gajewski, B. J. (2020) · BMC Medical Research Methodology · doi:10.1186/s12874-020-01097-6 · Cited by 5
Dynamic Representation in the American States, 1960–2012Preprint
Caughey, D., & Warshaw, C. (2014) · SSRN Electronic Journal · doi:10.2139/ssrn.2455441 · Cited by 8
Priors: Silent or Active Partners of Bayesian Inference?
Low‐Choy, S. (2012) · Wiley series in probability and statistics · doi:10.1002/9781118394472.ch3 · Cited by 12
Reusens, P., & Croux, C. (2015) · SSRN Electronic Journal · doi:10.2139/ssrn.2548960 · Cited by 5
Financial Regulatory Transparency and Sovereign Borrowing CostsPreprint
Copelovitch, M. G., Gandrud, C., & Hallerberg, M. (2015) · SSRN Electronic Journal · doi:10.2139/ssrn.2701852 · Cited by 5
Fast and flexible linear mixed models for genome-wide geneticsPreprint
Runcie, D. E., & Crawford, L. A. (2018) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/373902 · Cited by 5
Bayesian Function-on-Scalars Regression for High-Dimensional Data
Kowal, D. R., & Bourgeois, D. (2020) · Figshare · doi:10.6084/m9.figshare.11555739 · Cited by 5
Population-level inference for home-range areasPreprint
Fleming, C. H., Deznabi, I., Alavi, S., et al. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.07.05.451204 · Cited by 5
“Finally, our Bayesian analysis produced much larger small-m biases that we anticipated, even though we supplied a non-informative prior similar to that suggested by Gelman (2006) for variance parameters, and further assisted our Bayesian analysis by centering each prior on the truth.”
MethodAng, A., Chen, L., Gates, M. D., & Henderson, P. D. (2021) · Financial Analysts Journal · doi:10.1080/0015198x.2021.1960782 · Cited by 5
Low-dimensional encoding of decisions in parietal cortex reflects long-term training historyPreprint
Latimer, K. W., & Freedman, D. J. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.10.07.463576 · Cited by 5
Lee, S. (2022) · Mathematics · doi:10.20944/preprints202201.0407.v1 · Cited by 24
“See discussion by [95] for the prior options implemented on 8-schools example.”
MethodBayesian Approaches to Shrinkage and Sparse EstimationPreprint
Korobilis, D., & Shimizu, K. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3971832 · Cited by 5
Williams, D. R., Rouder, J. N., & Rast, P. (2019) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/gwatq · Cited by 5
Sarma, A., & Kay, M. (2020) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/fcmun · Cited by 9
Bayesian Hierarchical Modeling: An Introduction and ReassessmentPreprint
Veenman, M., Stefan, A. M., & Haaf, J. M. (2022) · doi:10.31234/osf.io/pskvx · Cited by 4
Cholinergic modulation of reinforcement learning and prefrontal value computations under uncertaintyPreprint
Kurtenbach, H., Froböse, M. I., Ort, E., Bahners, B. H., Butz, M., Schnitzler, A., Hirschmann, J., & Jocham, G. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.09.20.614105 · Cited by 4
Chan, I., Duh, R., & Tan, H. (2022) · European Accounting Review · doi:10.1080/09638180.2022.2074493 · Cited by 5
Holistic Energy Awareness and Robustness for Intelligent Drones
Saxena, R. R., Pal, J., Iyengar, S., Chhaglani, B., Ghosh, A., Padmanabhan, V. N., & Venkata, P. T. (2024) · ACM Transactions on Sensor Networks · doi:10.1145/3641855 · Cited by 6
“The priors were set using the recommendations provided by Gelman et al. [Gelman et al. 2006].”
MethodWiśniowski, A., & Raymer, J. (2025) · European Journal of Population / Revue européenne de Démographie · doi:10.1007/s10680-025-09729-7 · Cited by 7
“…deviations, i.e.: The half-t prior for the standard deviation is considered a reasonable choice over conjugate gamma priors because it has a heavy tail and ensures that the data can (23) A 1 ∼ Normal 0, 2 A , A ∼ t + ( 0, 2.5, 0.5 ) , (24) RS ∼ Normal 0, 0.2 2 . dominate inference (Gelman, 2006).”
InfluentialSimulation in the ‘Blind’ mind: Examining unconscious mental imagery in aphantasia
Muraki, E. J., & Pexman, P. M. (2025) · Neuropsychologia · doi:10.1016/j.neuropsychologia.2025.109218 · Cited by 7
Münch, F. F., & Koch, T. (2025) · Behavior Research Methods · doi:10.3758/s13428-025-02781-5 · Cited by 5
“For scale parameters(i.e.,residualstandarddeviations),wechosehalf-Cauchy priors (Gelman, 2006) with location 0 and scale .”
Shinagawa, K., Tanaka, Y., Terasawa, Y., & Umeda, S. (2025) · Neuroscience of Consciousness · doi:10.1093/nc/niaf059 · Cited by 14
Railway track geometry degradation: A review of prediction methods
Tarbozagh, A. S., Mohammadzadeh, S., Esmaeili, M., & Nadermohammady, S. (2026) · Results in Engineering · doi:10.1016/j.rineng.2026.109583 · Cited by 7
Predicting Field Degradation of Sealants Using Accelerated Tests from the NIST Solar Sphere
Pintar, A. L., White, C. C., Hunston, D. L., et al. (2017) · Elsevier eBooks · doi:10.1016/b978-0-323-49776-3.00009-x · Cited by 4
Berg, P., & Popescu, G. V. (2023) · Molecular & Cellular Proteomics · doi:10.1016/j.mcpro.2023.100658 · Cited by 4
“We use the expanded noncentered parameterization which has been shown to increase sampling convergence and ef fi ciency, and allows for increased model fl exibility (29 – 35).”
MethodAnalyzing population-level trials as N-of-1 trials: An application to gait
Zhou, L. J., Schneider, J., Bert Arnrich, A., & Konigorski, S. (2024) · Contemporary Clinical Trials Communications · doi:10.1016/j.conctc.2024.101282 · Cited by 6
“We assumed a half-Cauchy distribution for the residual variance 𝜎 (cf. Eq.”
Thermal and cardiovascular heat adaptations in active adolescents following summer
Brown, H. A., Topham, T. H., Clark, B., Woodward, A. P., Ioannou, L. G., Flouris, A. D., Telford, R. D., Smallcombe, J. W., Jay, O., & Périard, J. D. (2024) · Temperature · doi:10.1080/23328940.2024.2347161 · Cited by 6
“Flat priors were selected for the smooth terms [41], whilst a scaled prior was used on continuous variables [42].”
MethodSmit, R., Goodman, D. H., Boyce, J., et al. (2024) · River Research and Applications · doi:10.1002/rra.4341 · Cited by 4
Bayesian meta-analysis for rare outcomes
Alqasem, O. A., Demirhan, H., & Dolgun, A. (2025) · Journal of Biopharmaceutical Statistics · doi:10.1080/10543406.2025.2512205 · Cited by 4
Datta, G. S., Li, J., Yi, H., et al. (2025) · Journal of Survey Statistics and Methodology · doi:10.1093/jssam/smaf013 · Cited by 4
Zhang, J., Zhang, F., Tao, J., Yu, M. T. X., & Wen, F. (2025) · BMC Sports Science Medicine and Rehabilitation · doi:10.1186/s13102-025-01314-z · Cited by 6
Hedging sanctions risk: Cryptocurrency in central bank reserves
Ferranti, M. (2025) · Journal of International Money and Finance · doi:10.1016/j.jimonfin.2025.103433 · Cited by 23
Nagel, J. B., & Sudret, B. (2014) · Repository for Publications and Research Data (ETH Zurich) · doi:10.3929/ethz-a-010078997 · Cited by 4
Gajewski, B. J., & Dunton, N. E. (2012) · Medical Decision Making · doi:10.1177/0272989x12461855 · Cited by 4
A Noncentral t Regression Model for Meta-Analysis
Camilli, G., TORRE, J. D. L., & Chiu, C.-Y. (2010) · Journal of Educational and Behavioral Statistics · doi:10.3102/1076998609346966 · Cited by 4
Pragman, A. A., Isaacson, R. E., Wendt, C. H., & Reilly, C. (2015) · Journal of Computational Biology · doi:10.1089/cmb.2015.0021 · Cited by 4
Bayesian varying coefficient mixed-effects joint models with asymmetry and missingness
Lu, T., Cai, C. C., Lu, M., et al. (2017) · Statistical Modelling · doi:10.1177/1471082x16678543 · Cited by 4
Using phylogenetic comparative methods to gain insight into the evolution of social complexity
Fuentes-González, J. A., & Martins, E. P. (2019) · Behavioral Ecology and Sociobiology · doi:10.1007/s00265-018-2614-3 · Cited by 5
“For evolutionary rates, we also used diffuse priors through uniform distributions on the inverse of variance parameters: γ−1~U (0, 100), reflecting the common practice in Bayesian statistics of parameterizing the variance in terms of precision (Gelman 2006).”
MethodYe, S., Liang, Y., & Zhang, B. (2020) · Current Bioinformatics · doi:10.2174/1574893615999200520082636 · Cited by 6
Empirical Bayes method using surrounding pixel information for number and brightness analysis
Fukushima, R., Yamamoto, J., & Kinjo, M. (2021) · Biophysical Journal · doi:10.1016/j.bpj.2021.03.033 · Cited by 4
Hörberg, T., & Jaeger, T. F. (2021) · Frontiers in Psychology · doi:10.3389/fpsyg.2021.674202 · Cited by 5
“We used somewhat weaker priors than is standardly recommended for data analysis (e.g., Gelman, 2006; Gelman et al., 2008).”
MethodFox, J., & Smink, W. A. C. (2021) · Psychological Methods · doi:10.1037/met0000442 · Cited by 4
Unique population dynamics of Japanese field vole: Winter breeding and summer population decline
Murano, C., Iijima, H., & Azuma, N. (2022) · Population Ecology · doi:10.1002/1438-390x.12113 · Cited by 4
Mayer, D. J., & Fischer, R. (2022) · Social Science Quarterly · doi:10.1111/ssqu.13190 · Cited by 4
Parameterizing Lognormal state space models using moment matching
Smith, J. W., Thomas, R. Q., & Johnson, L. R. (2023) · Environmental and Ecological Statistics · doi:10.1007/s10651-023-00570-x · Cited by 7
“Following the advice of Gelman (2006) and Polson and Scott (2012), we use a central Half-Cauchy prior distribution for the precision parameters.”
MethodInfluentialEvidence of the cooperative response of Blattella germanica gut microbiota to antibiotic treatment
Creus-Martí, I., Marín‐Miret, J., Moyá, A., & Gómez, F. J. S. (2023) · Mathematical Biosciences · doi:10.1016/j.mbs.2023.109057 · Cited by 4
The neural dynamics of loss aversion
Nie, S., Wang, M., Li, J., Luo, H., & Zhang, H. (2023) · Imaging Neuroscience · doi:10.1162/imag_a_00047 · Cited by 5
“For the dynamic context model, we followed the same prior ( Gelman, 2006), and bounded the weights from 0 to 5 using an inverse-probit transformation: Here, Φ ( ⋅ ) is the cumulative distribution function of a standard normal distribution.”
Bayesian generalized additive model selection including a fast variational option
He, V. X., & WAND, M. P. (2023) · AStA Advances in Statistical Analysis · doi:10.1007/s10182-023-00490-y · Cited by 4
König, C., & Alexandrowicz, R. W. (2024) · Applied Psychological Measurement · doi:10.1177/01466216241227547 · Cited by 6
Adaptive Bayesian information borrowing methods for finding and optimizing subgroup-specific doses
Zhang, J., Lin, R., Chen, X., & Yan, F. (2024) · Clinical Trials · doi:10.1177/17407745231212193 · Cited by 7
“30 Second, due to the sparsity of data observed in early-phase trials, the performance of BHM can be sensitive to the priors, particularly those related to the inter-subgroup variance.”
Jaya, I. G. N. M., Kristiani, F., Andriyana, Y., & Chadidjah, A. (2024) · Mathematics · doi:10.3390/math12030451 · Cited by 4
Recommendation of Tahiti acid lime cultivars through Bayesian probability models
Malikouski, R. G., Ferreira, F. M., Chaves, S. F. D. S., Couto, E. G. D. O., Dias, K. O. D. G., & Bhering, L. L. (2024) · PLoS ONE · doi:10.1371/journal.pone.0299290 · Cited by 5
The R Package tipsae: Tools for Mapping Proportions and Indicators on the Unit Interval
Nicolò, S. D., & Gardini, A. (2024) · Journal of Statistical Software · doi:10.18637/jss.v108.i01 · Cited by 7
Peterson, E. N., Nethery, R. C., Padellini, T., Chen, J. T., Coull, B. A., Piel, F. B., Wakefield, J. C., Blangiardo, M., & Waller, L. A. (2024) · The Annals of Applied Statistics · doi:10.1214/23-aoas1849 · Cited by 6
Purg, N., Kraljič, A., Rahmati, M., Cho, Y., Ozimič, A. S., Murray, J. D., Anticevic, A., & Repovš, G. (2024) · Cerebral Cortex · doi:10.1093/cercor/bhae350 · Cited by 9
Redefining Item Response Models for Small Samples
Fox, J. (2024) · Journal of Educational and Behavioral Statistics · doi:10.3102/10769986241269886 · Cited by 8
Kürüm, E., Nguyen, D. V., Qian, Q., et al. (2024) · Lifetime Data Analysis · doi:10.1007/s10985-024-09635-w · Cited by 4
“…of the function that is approximated via the thin-plate approach and it is highly recommended to inspect the results under a number of different choices since the estimates of the variance components are known to be sensitive to the prior specification (Crainiceanu et al. 2005; Gelman 2006).”
BackgroundOhigashi, T., Maruo, K., Sozu, T., Sawamoto, R., & Gosho, M. (2024) · Pharmaceutical Statistics · doi:10.1002/pst.2453 · Cited by 7
“The base distribution, G 0 , must be carefully selected; for example, potential candidates include half-normal, half-t , log-normal and inverse-gamma distributions [6, 15, 22, 23].”
Frankemölle, J. P. K. W., Camps, J., Meutter, P. D., & Meyers, J. (2025) · Geoscientific model development · doi:10.5194/gmd-18-1989-2025 · Cited by 3
“The least informative prior with support on [0 , ∞ ) and given mean is the exponential distribution For f ( S ) , we choose the half-normal distribution (Gelman, 2006) Finally, we formulate f ( R ) .”
BackgroundSaville, B. R., Herring, A. H., & Kaufman, J. S. (2011) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/j.1467-985x.2011.00685.x · Cited by 6
“The continuous variables Gesti, NDIi, and Wtgaini are centered and standardized by 2 standard deviations to place the regression coefficients on a similar scale as the binary indicators (Gelman (2008)).”
BackgroundMethodRelating Demographic Characteristics of a Small Mammal to Remotely Sensed Forest-Stand Condition
Lada, H., Thomson, J. R., Cunningham, S. C., & Nally, R. M. (2014) · PLoS ONE · doi:10.1371/journal.pone.0091731 · Cited by 5
Pedroni, N., Zio, E., Pasanisi, A., et al. (2015) · ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems Part A Civil Engineering · doi:10.1061/ajrua6.0000848 · Cited by 4
Bayesian hierarchical duration model for repeated events: an application to behavioral observations
Dagne, G. A., & Snyder, J. J. (2009) · Journal of Applied Statistics · doi:10.1080/02664760802587032 · Cited by 5
A Bayesian Approach to Successive Comparisons
Aghamohammadi, A., Meshkani, M. R., & Mohammadzadeh, M. (2021) · Journal of Data Science · doi:10.6339/jds.2010.08(4).630 · Cited by 4
“Gelman (2006) investigate various priors for variance parameters and recommended a uniform distribution on the standard deviation.”
A comparison of alternative random regression and reaction norm models for whole genome predictions1
Yang, W., Chen, C., Steibel, J. P., Ernst, C. W., Bates, R. O., Zhou, L., & Tempelman, R. J. (2015) · Journal of Animal Science · doi:10.2527/jas.2014-8685 · Cited by 4
Gomes, M., Hatfield, L. A., & Normand, S. T. (2016) · Statistics in Medicine · doi:10.1002/sim.6969 · Cited by 3
Sub-optimality of some continuous shrinkage priors
Bhattacharya, A., Dunson, D. B., Pati, D., et al. (2016) · Stochastic Processes and their Applications · doi:10.1016/j.spa.2016.08.007 · Cited by 4
“In [24], the authors instead recommended a half-Cauchy prior as a default choice for the global variance (also see [23]).”
BackgroundResultInfluentialZero-Inflated Spatial Models: Application and Interpretation
Ainsworth, L. M., Dean, C. B., & Joy, R. (2016) · Lecture notes in statistics · doi:10.1007/978-3-319-31260-6_3 · Cited by 4
Hofer, J. D., & Rauk, A. P. (2016) · Journal of Pharmaceutical Sciences · doi:10.1016/j.xphs.2016.10.003 · Cited by 4
Mamouridis, V., Klein, N., Kneib, T., Cadarso-Suárez, C., & Maynou, F. (2016) · Mathematical Biosciences · doi:10.1016/j.mbs.2016.11.016 · Cited by 5
Derrick, S. M., Hipp, J. A., & Goel, P. (2017) · Elsevier eBooks · doi:10.1016/b978-0-12-805429-1.00023-5 · Cited by 6
Castaño, L. S. Á., Caicedo-Velásquez, B., Castaño-Díez, C., Marí-Dell’Olmo, M., & Gotsens, M. (2018) · Biomédica · doi:10.7705/biomedica.v38i3.3734 · Cited by 4
“0001 was assigned to the standard deviations σv and σu (21).”
MethodChin, V., Gunawan, D., Fiebig, D. G., Kohn, R., & Sisson, S. A. (2020) · Journal of the Royal Statistical Society Series C (Applied Statistics) · doi:10.1111/rssc.12393 · Cited by 6
“The marginal prior of the standard deviation in Σα is a half-t(λ,Ai) distribution, as suggested in Gelman (2006).”
BackgroundAccounting for Data Architecture on Structural Equation Modeling of Feedlot Cattle Performance
Chitakasempornkul, K., Sanderson, M. W., Cha, E., Renter, D. G., Jager, A., & Bello, N. M. (2018) · Journal of Agricultural Biological and Environmental Statistics · doi:10.1007/s13253-018-0336-7 · Cited by 4
Forecasting of Used Product Returns for Remanufacturing
Geda, M. W., & Kwong, C. K. (2018) · IEEE International Conference on Industrial Engineering and Engineering Management (IEEM) · doi:10.1109/ieem.2018.8607362 · Cited by 7
Bakar, K. S., Biddle, N., Kokic, P. N., et al. (2019) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1111/rssa.12526 · Cited by 4
“…of the inverse gamma prior distribution be a =aζ =aξ =aν = aφ =2 and b =bζ =bξ =bν =bφ =1, as the inverse-gamma.2, 1/ distribution is often treated as proper and weakly informative (see the details in Gelman (2006), Gelman et al. (2008), Bakar and Kokic (2017) and Sahu and Bakar (2012b)).”
BackgroundBayesian inference of natural selection from spatiotemporal phenotypic data
David, O., van Frank, G., Goldringer, I., et al. (2019) · Theoretical Population Biology · doi:10.1016/j.tpb.2019.11.007 · Cited by 4
Kozono, J., Nishihara, G. N., Endo, H., & Terada, R. (2020) · Journal of Applied Phycology · doi:10.1007/s10811-019-01967-7 · Cited by 4
“Informative normal priors were placed on all parameters of the model by using values from a previous study (Kokubu et al. 2015), and a half-Cauchy prior distribution was placed on the scale parameter of the models (Gelman 2004, 2006).”
MethodSuzuki, S., & Yamada, T. (2020) · PeerJ · doi:10.7717/peerj.8722 · Cited by 5
“For the concentration parameters, we set the half228 Student’s t-distribution as a prior distribution (Gelman, 2006).”
MethodHierarchical multidimensional scaling for the comparison of musical performance styles
Yanchenko, A. K., & Hoff, P. D. (2020) · The Annals of Applied Statistics · doi:10.1214/20-aoas1391 · Cited by 4
Incorporating Design Weights And Historical Data Into Model-Based Small-Area Estimation
Xie, H. W., Barker, L. E., & Rolka, D. B. (2021) · Journal of Data Science · doi:10.6339/jds.202001_18(1).0006 · Cited by 4
Montoya-Betancur, K. V., Caicedo-Velásquez, B., & Castaño, L. S. Á. (2020) · Cadernos de Saúde Pública · doi:10.1590/0102-311x00101219 · Cited by 6
Risk factor identification in cystic fibrosis by flexible hierarchical joint models
Su, W., Wang, X., & Szczesniak, R. D. (2020) · Statistical Methods in Medical Research · doi:10.1177/0962280220950369 · Cited by 4
on behalf of the PERC-KIDSCAN Ketodex Study Group, Heath, A., Ríos, J. D., Pullenayegum, E., Pechlivanoglou, P., Offringa, M., Yaskina, M., Watts, R., Rimmer, S., Klassen, T. P., et al. (2021) · Trials · doi:10.1186/s13063-020-04946-3 · Cited by 6
“Priors for the standard deviations will use non-central t distributions (truncated at 0) [28].”
MethodGonzales, K. L., Famula, T. R., Feng, L. C., Power, H. M. N., & Bullis, J. M. (2021) · Journal of Small Animal Practice · doi:10.1111/jsap.13292 · Cited by 5
Skew normal small area time models for the Brazilian annual service sector survey
Neves, A. F., Silva, D. B. D. N., & Moura, F. A. S. (2020) · Statistics in Transition New Series · doi:10.21307/stattrans-2020-032 · Cited by 4
“As discussed in Gelman (2006), the use of this prior guarantees a proper posterior density as well as other desirable properties.”
BackgroundMethod
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.