BAYESIAN MODEL ASSESSMENT IN FACTOR ANALYSIS

Factor analysis has been one of the most powerful and flexible tools for assessment of multivariate dependence and codependence. Loosely speaking, it could be argued that the origin of its success rests in its very exploratory nature, where various kinds of data-relationships amongst the variables at study can be iteratively verified and/or refuted. Bayesian inference in factor analytic models has received renewed attention in recent years, partly due to computational advances but also partly to applied focuses generating factor structures as exemplified by recent work in financial time series modeling. The focus of our current work is on exploring questions of uncertainty about the number of latent factors in a multi- variate factor model, combined with methodological and computational issues of model specification and model fitting. We explore reversible jump MCMC methods that build on sets of parallel Gibbs sampling-based analyses to generate suitable empirical proposal distributions and that address the challenging problem of finding efficient proposals in high-dimensional models. Alternative MCMC methods based on bridge sampling are discussed, and these fully Bayesian MCMC approaches are compared with a collection of popular model selection methods in empirical stud- ies. Various additional computational issues are discussed, including situations where prior information is scarce, and the methods are explored in studies of some simulated data sets and an econometric time series example.

[1]  A. Basilevsky,et al.  Factor Analysis as a Statistical Method. , 1964 .

[2]  K. Jöreskog.  Some contributions to maximum likelihood factor analysis , 1967 .

[3]  G. C. Tiao,et al.  Bayesian inference in statistical analysis , 1973 .

[4]  Arnold Zellner,et al.  An Introduction to Bayesian Inference in Econometrics. , 1974 .

[5]  R. P. McDonald,et al.  Bayesian estimation in unrestricted factor analysis: A treatment for heywood cases , 1975 .

[6]  J. Geweke,et al.  Interpreting the Likelihood Ratio Statistic in Factor Models When Sample Size is Small , 1980 .

[7]  D. Bartholomew.  Posterior analysis of the factor model , 1981 .

[8]  Sik-Yum Lee.  A bayesian approach to confirmatory factor analysis , 1981 .

[9]  A. Dawid.  Some matrix-variate distribution theory: Notational considerations and a Bayesian application , 1981 .

[10]  D. Spiegelhalter,et al.  Bayes Factors for Linear and Log‐Linear Models with Vague Prior Information , 1982 .

[11]  S. J. Press,et al.  Applied multivariate analysis : using Bayesian and frequentist methods of inference , 1984 .

[12]  L. Tierney,et al.  Accurate Approximations for Posterior Moments and Marginal Densities , 1986 .

[13]  Donald E. Ramirez,et al.  An Expert Model Selection Approach to Determine the “Best” Pattern Structure in Factor Analysis Models , 1987 .

[14]  H. Akaike.  Factor analysis and AIC , 1987 .

[15]  J. Besag.  A candidate's formula: A curious result in Bayesian prediction , 1989 .

[16]  M. West,et al.  Bayesian forecasting and dynamic models , 1989 .

[17]  S. J. Press,et al.  Bayesian Inference in Factor Analysis , 1989 .

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

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

[20]  A. O'Hagan,et al.  Fractional Bayes factors for model comparison , 1995 .

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

[22]  J. Geweke,et al.  Measuring the pricing error of the arbitrage pricing theory , 1996 .

[23]  Y. Kano,et al.  Identifiability of full, marginal, and conditional factor analysis models , 1995 .

[24]  S. Chib.  Marginal Likelihood from the Gibbs Output , 1995 .

[25]  J. Berger,et al.  The Intrinsic Bayes Factor for Model Selection and Prediction , 1996 .

[26]  Peter Green,et al.  Markov chain Monte Carlo in Practice , 1996 .

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

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

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

[30]  M. Pitt,et al.  Time Varying Covariances: A Factor Stochastic Volatility Approach (with discussion , 1998 .

[31]  H. Bozdogan,et al.  Bayesian Factor Analysis Model and Choosing the Number of Factors Using a New Informational Complexity Criterion , 1998 .

[32]  B. Muthén,et al.  A Bayesian approach to nonlinear latent variable models using the Gibbs sampler and the metropolis-hastings algorithm , 1998 .

[33]  Alan E. Gelfand,et al.  Model choice: A minimum posterior predictive loss approach , 1998, AISTATS.

[34]  Wolfgang Polasek,et al.  Factor analysis and outliers: A Bayesian approach , 1999 .

[35]  P. Green,et al.  Trans-dimensional Markov chain Monte Carlo , 2000 .

[36]  M. West,et al.  Bayesian dynamic factor models and variance matrix dis-counting for portfolio allocation , 2000 .

[37]  S. Godsill.  On the Relationship Between Markov chain Monte Carlo Methods for Model Uncertainty , 2001 .

[38]  Stan Lipovetsky,et al.  Latent Variable Models and Factor Analysis , 2001, Technometrics.

[39]  Daniel Gianola,et al.  Bayesian Analysis of Linear Models , 2002 .

[40]  Petros Dellaportas,et al.  On Bayesian model and variable selection using MCMC , 2002, Stat. Comput..

[41]  J. Berger,et al.  Expected‐posterior prior distributions for model selection , 2002 .

[42]  H. Migon,et al.  Comovements and contagion in emergent markets: stock indexes volatilities , 2002 .

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

[44]  P. Müller,et al.  Bayesian Meta‐analysis for Longitudinal Data Models Using Multivariate Mixture Priors , 2003, Biometrics.

[45]  S. James Press,et al.  Subjective and objective Bayesian statistics : principles, models, and applications , 2003 .

[46]  O. Cappé,et al.  Reversible jump, birth‐and‐death and more general continuous time Markov chain Monte Carlo samplers , 2003 .