The maximum approximate composite marginal likelihood (MACML) estimation of multinomial probit-based unordered response choice models

Abstract The likelihood functions of multinomial probit (MNP)-based choice models entail the evaluation of analytically-intractable integrals. As a result, such models are usually estimated using maximum simulated likelihood (MSL) techniques. Unfortunately, for many practical situations, the computational cost to ensure good asymptotic MSL estimator properties can be prohibitive and practically infeasible as the number of dimensions of integration rises. In this paper, we introduce a maximum approximate composite marginal likelihood (MACML) estimation approach for MNP models that can be applied using simple optimization software for likelihood estimation. It also represents a conceptually and pedagogically simpler procedure relative to simulation techniques, and has the advantage of substantial computational time efficiency relative to the MSL approach. The paper provides a “blueprint” for the MACML estimation for a wide variety of MNP models.

[1]  D. Louis,et al.  A pseudolikelihood approach for simultaneous analysis of array comparative genomic hybridizations. , 2005, Biostatistics.

[2]  Richard L. Smith,et al.  Asymptotic properties of computationally efficient alternative estimators for a class of multivariate normal models , 2007 .

[3]  Peter S. Craig,et al.  A new reconstruction of multivariate normal orthant probabilities , 2007 .

[4]  Denzil G. Fiebig,et al.  The Generalized Multinomial Logit Model: Accounting for Scale and Coefficient Heterogeneity , 2010, Mark. Sci..

[5]  Varameth Vichiensan,et al.  Discrete Choice Model with Structuralized Spatial Effects for Location Analysis , 2004 .

[6]  C. Winston,et al.  UNCOVERING THE DISTRIBUTION OF MOTORISTS' PREFERENCES FOR TRAVEL TIME AND RELIABILITY : IMPLICATIONS FOR ROAD PRICING , 2002 .

[7]  Luc Anselin,et al.  Spatial Externalities, Spatial Multipliers, And Spatial Econometrics , 2003 .

[8]  N. Reid,et al.  AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS , 2011 .

[9]  James J. Heckman,et al.  ECONOMETRIC ANALYSIS OF LONGITUDINAL DATA , 1986 .

[10]  H. Joe,et al.  Composite likelihood estimation in multivariate data analysis , 2005 .

[11]  A. Holly,et al.  Estimation of multivariate probit models by exact maximum likelihood , 2009 .

[12]  Harry Joe,et al.  On weighting of bivariate margins in pairwise likelihood , 2009, J. Multivar. Anal..

[13]  K. Train.  Discrete Choice Methods with Simulation , 2003 .

[14]  C. Bhat.  The MACML Estimation of the Normally-Mixed Multinomial Logit Model , 2011 .

[15]  D. McFadden,et al.  MIXED MNL MODELS FOR DISCRETE RESPONSE , 2000 .

[16]  Nils Lid Hjort,et al.  ML, PL, QL in Markov Chain Models , 2008 .

[17]  G. Leckie.  Bulletin of the International Statistical Institute , 2013 .

[18]  C. Bhat.  Simulation estimation of mixed discrete choice models using randomized and scrambled Halton sequences , 2003 .

[19]  Luc Anselin,et al.  New Directions in Spatial Econometrics , 2011 .

[20]  Mark M. Fleming.  Techniques for Estimating Spatially Dependent Discrete Choice Models , 2004 .

[21]  Chandra R. Bhat,et al.  A MIXED SPATIALLY CORRELATED LOGIT MODEL: FORMULATION AND APPLICATION TO RESIDENTIAL CHOICE MODELING , 2004 .

[22]  P. Switzer,et al.  Estimation of spatial distributions from point sources with application to air pollution measurement. Technical report No. 9 , 1977 .

[23]  Chandra R. Bhat,et al.  A UNIFIED MIXED LOGIT FRAMEWORK FOR MODELING REVEALED AND STATED PREFERENCES: FORMULATION AND APPLICATION TO CONGESTION PRICING ANALYSIS IN THE SAN FRANCISCO BAY AREA , 2002 .

[24]  V. P. Godambe.  An Optimum Property of Regular Maximum Likelihood Estimation , 1960 .

[25]  Carlo Gaetan,et al.  Composite likelihood methods for space-time data , 2006 .

[26]  D. Cox,et al.  A note on pseudolikelihood constructed from marginal densities , 2004 .

[27]  David J. Nott,et al.  A pairwise likelihood approach to analyzing correlated binary data , 2000 .

[28]  Chandra R. Bhat,et al.  The Impact of Stop-Making and Travel Time Reliability on Commute Mode Choice , 2006 .

[29]  Chandra R. Bhat,et al.  Flexible Model Structures for Discrete Choice Analysis , 2007 .

[30]  Chandra R. Bhat,et al.  A Multivariate Ordered-response Model System for Adults' Weekday Activity Episode Generation by Activity Purpose and Social Context , 2010 .

[31]  Denis Bolduc,et al.  The Effect of Incentive Policies on the Practice Location of Doctors: A Multinomial Probit Analysis , 1996, Journal of Labor Economics.

[32]  Florian Heiss,et al.  Likelihood approximation by numerical integration on sparse grids , 2008 .

[33]  A. Genz,et al.  Numerical computation of multivariate t-probabilities with application to power calculation of multiple contrasts , 1999 .

[34]  C. Varin,et al.  A comparison of the maximum simulated likelihood and composite marginal likelihood estimation approaches in the context of the multivariate ordered-response model , 2010 .

[35]  J. LeSage.  Bayesian Estimation of Limited Dependent Variable Spatial Autoregressive Models , 2010 .

[36]  C. Varin.  On composite marginal likelihoods , 2008 .

[37]  Cristiano Varin,et al.  Pairwise Likelihood Inference for General State Space Models , 2008 .

[38]  Joris Pinkse,et al.  Contracting in space: An application of spatial statistics to discrete-choice models , 1998 .

[39]  Chandra R. Bhat,et al.  A copula-based closed-form binary logit choice model for accommodating spatial correlation across observational units , 2009, J. Geogr. Syst..

[40]  Roberto S. Mariano,et al.  Simulation-based Inference in Econometrics , 2008 .

[41]  P. Zarembka.  Frontiers in econometrics , 1973 .

[42]  Chandra R. Bhat,et al.  A simulation evaluation of the maximum approximate composite marginal likelihood (MACML) estimator for mixed multinomial probit models , 2011 .

[43]  Robert F. Engle,et al.  Fitting and Testing Vast Dimensional Time-Varying Covariance Models , 2007 .

[44]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[45]  C. Varin,et al.  A mixed autoregressive probit model for ordinal longitudinal data. , 2010, Biostatistics.

[46]  Luigi Pace,et al.  ADJUSTING COMPOSITE LIKELIHOOD RATIO STATISTICS , 2009 .

[47]  Geert Molenberghs,et al.  A pairwise likelihood approach to estimation in multilevel probit models , 2004, Comput. Stat. Data Anal..

[48]  Andrew K. Rose,et al.  Empirical research on nominal exchange rates , 1995 .

[49]  Cristiano Varin,et al.  Pairwise likelihood inference for ordinal categorical time series , 2006, Comput. Stat. Data Anal..

[50]  Mohand L. Feddag,et al.  Pairwise likelihood for the longitudinal mixed Rasch model , 2009, Comput. Stat. Data Anal..

[51]  Wim P. M. Vijverberg,et al.  Why Cooperate? Public Goods, Economic Power, and the Montreal Protocol , 2003, Review of Economics and Statistics.

[52]  D. Hensher,et al.  Willingness to pay for travel time reliability in passenger transport: A review and some new empirical evidence , 2010 .

[53]  Chandra R. Bhat,et al.  Modeling the choice continuum: an integrated model of residential location, auto ownership, bicycle ownership, and commute tour mode choice decisions , 2011 .

[54]  G. Molenberghs,et al.  Models for Discrete Longitudinal Data , 2005 .

[55]  Subhash R Lele,et al.  Sampling variability and estimates of density dependence: a composite-likelihood approach. , 2006, Ecology.

[56]  Peter E. Rossi,et al.  Bayesian analysis of the multinomial probit model , 2000 .

[57]  Bart J. Bronnenberg,et al.  Structural Applications of the Discrete Choice Model , 2002 .

[58]  C. Bhat.  Quasi-random maximum simulated likelihood estimation of the mixed multinomial logit model , 2001 .

[59]  Robert J. Franzese,et al.  Empirical Models of Spatial Interdependence , 2008 .

[60]  H. Joe.  Approximations to Multivariate Normal Rectangle Probabilities Based on Conditional Expectations , 1995 .

[61]  Estimating Mixtures of Discrete Choice Model , 2007 .

[62]  Wim P. M. Vijverberg,et al.  Probit in a Spatial Context: A Monte Carlo Analysis , 2004 .

[63]  David Ruppert,et al.  Aberrant Crypt Foci and Semiparametric Modeling of Correlated Binary Data , 2008, Biometrics.

[64]  C. Varin,et al.  A note on composite likelihood inference and model selection , 2005 .

[65]  Chandra R. Bhat,et al.  Accommodating Spatial Correlation Across Choice Alternatives in Discrete Choice Models: Application to Modeling Residential Location Choice Behavior , 2011 .

[66]  J. C. van Houwelingen,et al.  Logistic Regression for Correlated Binary Data , 1994 .

[67]  C. Bhat,et al.  A Flexible Spatially Dependent Discrete Choice Model: Formulation and Application to Teenagers’ Weekday Recreational Activity Participation , 2010 .

[68]  Henry E. Brady,et al.  The Oxford Handbook of Political Methodology , 2010 .

[69]  Tamás Szántai,et al.  Computing Multivariate Normal Probabilities: A New Look , 2002 .

[70]  Harry Joe,et al.  Accuracy of Laplace approximation for discrete response mixed models , 2008, Comput. Stat. Data Anal..

[71]  John M. Rose,et al.  Allowing for intra-respondent variations in coefficients estimated on repeated choice data , 2009 .

[72]  K. Mardia,et al.  Maximum likelihood estimation using composite likelihoods for closed exponential families , 2009 .

[73]  Denzil G. Fiebig,et al.  Consumers and experts: an econometric analysis of the demand for water heaters , 2006 .

[74]  Florian Heiss,et al.  The panel probit model: Adaptive integration on sparse grids , 2010 .

[75]  D. McFadden.  Conditional logit analysis of qualitative choice behavior , 1972 .

[76]  J. Kent.  Robust properties of likelihood ratio tests , 1982 .

[77]  Andrew R. Solow,et al.  A method for approximating multivariate normal orthant probabilities , 1990 .

[78]  Sergio J. Rey,et al.  Advances in Spatial Econometrics: Methodology, Tools and Applications , 2004 .

[79]  Patrick J. Heagerty,et al.  Window Subsampling of Estimating Functions with Application to Regression Models , 2000 .