A tutorial on bridge sampling

The marginal likelihood plays an important role in many areas of Bayesian statistics such as parameter estimation, model comparison, and model averaging. In most applications, however, the marginal likelihood is not analytically tractable and must be approximated using numerical methods. Here we provide a tutorial on bridge sampling (Bennett, 1976; Meng & Wong, 1996), a reliable and relatively straightforward sampling method that allows researchers to obtain the marginal likelihood for models of varying complexity. First, we introduce bridge sampling and three related sampling methods using the beta-binomial model as a running example. We then apply bridge sampling to estimate the marginal likelihood for the Expectancy Valence (EV) model—a popular model for reinforcement learning. Our results indicate that bridge sampling provides accurate estimates for both a single participant and a hierarchical version of the EV model. We conclude that bridge sampling is an attractive method for mathematical psychologists who typically aim to approximate the marginal likelihood for a limited set of possibly high-dimensional models.

[1]  Xiao-Li Meng,et al.  Simulating Normalizing Constants: From Importance Sampling to Bridge Sampling to Path Sampling , 1998 .

[2]  J. Dickey,et al.  The Weighted Likelihood Ratio, Sharp Hypotheses about Chances, the Order of a Markov Chain , 1970 .

[3]  Radford M. Neal.  Annealed importance sampling , 1998, Stat. Comput..

[4]  Eric-Jan Wagenmakers,et al.  An encompassing prior generalization of the Savage-Dickey density ratio , 2010, Comput. Stat. Data Anal..

[5]  14 Model Comparison and the Principle of Parsimony , 2022 .

[6]  H. Damasio,et al.  Dissociation Of Working Memory from Decision Making within the Human Prefrontal Cortex , 1998, The Journal of Neuroscience.

[7]  T. Ando.  Bayesian predictive information criterion for the evaluation of hierarchical Bayesian and empirical Bayes models , 2007 .

[8]  D. Martino,et al.  Neuropsychological frontal impairments and negative symptoms in schizophrenia , 2007, Psychiatry Research.

[9]  Robin J. Prescott,et al.  Generalised Linear Mixed Models , 2006 .

[10]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[11]  E. Wagenmakers,et al.  Bayes factors for reinforcement-learning models of the Iowa Gambling Task , 2016 .

[12]  Helen Steingroever,et al.  Validating the PVL-Delta model for the Iowa gambling task , 2013, Front. Psychol..

[13]  Alexander Etz,et al.  J. B. S. Haldane's Contribution to the Bayes Factor Hypothesis Test , 2015, 1511.08180.

[14]  Dale J. Poirier,et al.  The Growth of Bayesian Methods in Statistics and Economics Since 1970 , 2006 .

[15]  A. Raftery,et al.  Stopping the Gibbs Sampler,the Use of Morphology,and Other Issues in Spatial Statistics (Bayesian image restoration,with two applications in spatial statistics) -- (Discussion) , 1991 .

[16]  L. Wasserman,et al.  Computing Bayes Factors by Combining Simulation and Asymptotic Approximations , 1997 .

[17]  Richard G. Everitt,et al.  Likelihood-free estimation of model evidence , 2011 .

[18]  R. Rescorla,et al.  A theory of Pavlovian conditioning : Variations in the effectiveness of reinforcement and nonreinforcement , 1972 .

[19]  Gregory P. Lee,et al.  Different Contributions of the Human Amygdala and Ventromedial Prefrontal Cortex to Decision-Making , 1999, The Journal of Neuroscience.

[20]  R. Blair,et al.  Somatic Markers and Response Reversal: Is There Orbitofrontal Cortex Dysfunction in Boys with Psychopathic Tendencies? , 2001, Journal of abnormal child psychology.

[21]  Jean-Paul Fox,et al.  Evaluating evidence for invariant items: A Bayes factor applied to testing measurement invariance in IRT models , 2016 .

[22]  R. Duncan Luce,et al.  Individual Choice Behavior , 1959 .

[23]  Laura Bellodi,et al.  Decision-making heterogeneity in obsessive-compulsive disorder: ventromedial prefrontal cortex function predicts different treatment outcomes , 2002, Neuropsychologia.

[24]  Tom Lodewyckx,et al.  A tutorial on Bayes factor estimation with the product space method , 2011 .

[25]  G. Nicholls,et al.  Bridge estimation of the probability density at a point , 2001 .

[26]  Adrian E. Raftery,et al.  Bayesian Model Averaging: A Tutorial , 2016 .

[27]  Jun Lu,et al.  Signal Detection Models with Random Participant and Item Effects , 2007 .

[28]  W. Michael Conklin,et al.  Monte Carlo Methods in Bayesian Computation , 2001, Technometrics.

[29]  M. Plummer,et al.  CODA: convergence diagnosis and output analysis for MCMC , 2006 .

[30]  E. Wagenmakers,et al.  Psychological interpretation of the ex-Gaussian and shifted Wald parameters: A diffusion model analysis , 2009, Psychonomic bulletin & review.

[31]  E. Ionides.  Truncated Importance Sampling , 2008 .

[32]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[33]  I. J. Myung,et al.  GUEST EDITORS' INTRODUCTION: Special Issue on Model Selection , 2000 .

[34]  S. E. Ahmed,et al.  Markov Chain Monte Carlo: Stochastic Simulation for Bayesian Inference , 2008, Technometrics.

[35]  Young K. Truong,et al.  Polynomial splines and their tensor products in extended linear modeling: 1994 Wald memorial lecture , 1997 .

[36]  Sanjay Jain,et al.  Editors' Introduction , 2005, ALT.

[37]  E. Wagenmakers,et al.  Bayesian hypothesis testing for psychologists: A tutorial on the Savage–Dickey method , 2010, Cognitive Psychology.

[38]  A. Damasio,et al.  Insensitivity to future consequences following damage to human prefrontal cortex , 1994, Cognition.

[39]  David E. Jones,et al.  Warp Bridge Sampling: The Next Generation , 2016, Journal of the American Statistical Association.

[40]  Eric-Jan Wagenmakers,et al.  A Comparison of Reinforcement Learning Models for the Iowa Gambling Task Using Parameter Space Partitioning , 2013, J. Probl. Solving.

[41]  Jeffrey N. Rouder,et al.  A hierarchical model for estimating response time distributions , 2005, Psychonomic bulletin & review.

[42]  A. Damasio,et al.  Deciding Advantageously Before Knowing the Advantageous Strategy , 1997, Science.

[43]  Woojae Kim,et al.  A Model-Based fMRI Analysis with Hierarchical Bayesian Parameter Estimation. , 2013, Journal of neuroscience, psychology, and economics.

[44]  T. Griffiths,et al.  Modeling individual differences using Dirichlet processes , 2006 .

[45]  Jun Lu,et al.  An introduction to Bayesian hierarchical models with an application in the theory of signal detection , 2005, Psychonomic bulletin & review.

[46]  Michael D. Lee,et al.  A Survey of Model Evaluation Approaches With a Tutorial on Hierarchical Bayesian Methods , 2008, Cogn. Sci..

[47]  G. Kesteven,et al.  The Coefficient of Variation , 1946, Nature.

[48]  Anna Pajor,et al.  Estimating the Marginal Likelihood Using the Arithmetic Mean Identity , 2017 .

[49]  H. Stern,et al.  An Empirical Comparison of Methods for Computing Bayes Factors in Generalized Linear Mixed Models , 2005 .

[50]  Adrian E. Raftery,et al.  Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors , 1999 .

[51]  Wolf Vanpaemel,et al.  Prototypes, exemplars and the response scaling parameter: A Bayes factor perspective , 2016 .

[52]  Christian P. Robert,et al.  The expected demise of the Bayes factor , 2015, 1506.08292.

[53]  David J. Lunn,et al.  Generic reversible jump MCMC using graphical models , 2009, Stat. Comput..

[54]  Thorsten Pachur,et al.  Bayesian techniques for analyzing group differences in the Iowa Gambling Task: A case study of intuitive and deliberate decision-makers , 2018, Psychonomic bulletin & review.

[55]  J. Dickey.  The Weighted Likelihood Ratio, Linear Hypotheses on Normal Location Parameters , 1971 .

[56]  B. Carlin,et al.  Bayesian Model Choice Via Markov Chain Monte Carlo Methods , 1995 .

[57]  Jonathan J. Forster,et al.  Default Bayesian model determination methods for generalised linear mixed models , 2010, Comput. Stat. Data Anal..

[58]  J. Busemeyer,et al.  A contribution of cognitive decision models to clinical assessment: decomposing performance on the Bechara gambling task. , 2002, Psychological assessment.

[59]  E. Wagenmakers,et al.  Bayesian parameter estimation in the Expectancy Valence model of the Iowa gambling task , 2010 .

[60]  Roberto Keller,et al.  Frontal lobe dysfunction in pathological gambling patients , 2002, Biological Psychiatry.

[61]  Darrell A. Worthy,et al.  A Comparison Model of Reinforcement-Learning and Win-Stay-Lose-Shift Decision-Making Processes: A Tribute to W.K. Estes. , 2014, Journal of mathematical psychology.

[62]  E. Wagenmakers,et al.  Model Comparison and the Principle of Parsimony , 2015 .

[63]  Xiao-Li Meng,et al.  Warp Bridge Sampling , 2002 .

[64]  M. Lindgren,et al.  Editors' introduction (Editorial) , 2015 .

[65]  A. Raftery,et al.  Estimating Bayes Factors via Posterior Simulation with the Laplace—Metropolis Estimator , 1997 .

[66]  Thom Baguley,et al.  Prior approval: the growth of Bayesian methods in psychology. , 2013, The British journal of mathematical and statistical psychology.

[67]  H. Damasio,et al.  Characterization of the decision-making deficit of patients with ventromedial prefrontal cortex lesions. , 2000, Brain : a journal of neurology.

[68]  E. Wagenmakers,et al.  Bayesian Estimation of Multinomial Processing Tree Models with Heterogeneity in Participants and Items , 2013, Psychometrika.

[69]  J. Busemeyer,et al.  An improved cognitive model of the Iowa and Soochow Gambling Tasks with regard to model fitting performance and tests of parameter consistency , 2015, Front. Psychol..

[70]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[71]  E. Wagenmakers,et al.  Harold Jeffreys’s default Bayes factor hypothesis tests: Explanation, extension, and application in psychology , 2016 .

[72]  Ioannis Ntzoufras,et al.  Bayesian Model and Variable Evaluation , 2008 .

[73]  S. Chib,et al.  Marginal Likelihood From the Metropolis–Hastings Output , 2001 .

[74]  Kaileigh A. Byrne,et al.  Decomposing the roles of perseveration and expected value representation in models of the Iowa gambling task , 2013, Front. Psychol..

[75]  I. J. Myung,et al.  Toward a method of selecting among computational models of cognition. , 2002, Psychological review.

[76]  Ruud Wetzels,et al.  A Bayesian test for the hot hand phenomenon , 2016 .

[77]  James O. Berger,et al.  Rejection odds and rejection ratios: A proposal for statistical practice in testing hypotheses , 2015, Journal of mathematical psychology.

[78]  L. M. M.-T..  Theory of Probability , 1929, Nature.

[79]  Jeffrey N. Rouder,et al.  A hierarchical process-dissociation model. , 2008, Journal of experimental psychology. General.

[80]  E. Wagenmakers,et al.  Absolute performance of reinforcement-learning models for the Iowa Gambling Task , 2014 .

[81]  Charles H. Bennett,et al.  Efficient estimation of free energy differences from Monte Carlo data , 1976 .

[82]  Thorsten Pachur,et al.  Using Bayesian hierarchical parameter estimation to assess the generalizability of cognitive models of choice , 2015, Psychonomic bulletin & review.

[83]  M. Newton.  Approximate Bayesian-inference With the Weighted Likelihood Bootstrap , 1994 .

[84]  Andrew Thomas,et al.  WinBUGS - A Bayesian modelling framework: Concepts, structure, and extensibility , 2000, Stat. Comput..

[85]  M. Lee.  Three case studies in the Bayesian analysis of cognitive models , 2008, Psychonomic bulletin & review.

[86]  A. Owen,et al.  Safe and Effective Importance Sampling , 2000 .

[87]  Rémi Bardenet,et al.  Monte Carlo Methods , 2013, Encyclopedia of Social Network Analysis and Mining. 2nd Ed..

[88]  Eric-Jan Wagenmakers,et al.  Editors’ introduction to the special issue “Bayes factors for testing hypotheses in psychological research : Practical relevance and new developments” , 2016 .

[89]  Charles E. Brown.  Applied Multivariate Statistics in Geohydrology and Related Sciences , 1998 .

[90]  James O. Berger,et al.  Posterior model probabilities via path‐based pairwise priors , 2005 .

[91]  B. Bogerts,et al.  Deficit in decision making in catatonic schizophrenia: An exploratory study , 2005, Psychiatry Research.

[92]  A. Gelfand,et al.  Bayesian Model Choice: Asymptotics and Exact Calculations , 1994 .

[93]  Martyn Plummer,et al.  JAGS: A program for analysis of Bayesian graphical models using Gibbs sampling , 2003 .

[94]  Xiao-Li Meng,et al.  SIMULATING RATIOS OF NORMALIZING CONSTANTS VIA A SIMPLE IDENTITY: A THEORETICAL EXPLORATION , 1996 .

[95]  Eric-Jan Wagenmakers,et al.  An evaluation of alternative methods for testing hypotheses, from the perspective of Harold Jeffreys , 2016 .

[96]  Jerome R. Busemeyer,et al.  Comparison of Decision Learning Models Using the Generalization Criterion Method , 2008, Cogn. Sci..

[97]  S. Frühwirth-Schnatter.  Estimating Marginal Likelihoods for Mixture and Markov Switching Models Using Bridge Sampling Techniques , 2004 .