A Bayesian Multinomial Probit MODEL FOR THE ANALYSIS OF PANEL CHOICE DATA

Duncan K.H. Fong, Sunghoon Kim, Zhe Chen, Wayne S. DeSarbo

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

A new Bayesian multinomial probit model is proposed for the analysis of panel choice data. Using a parameter expansion technique, we are able to devise a Markov Chain Monte Carlo algorithm to compute our Bayesian estimates efficiently. We also show that the proposed procedure enables the estimation of individual level coefficients for the single-period multinomial probit model even when the available prior information is vague. We apply our new procedure to consumer purchase data and reanalyze a well-known scanner panel dataset that reveals new substantive insights. In addition, we delineate a number of advantageous features of our proposed procedure over several benchmark models. Finally, through a simulation analysis employing a fractional factorial design, we demonstrate that the results from our proposed model are quite robust with respect to differing factors across various conditions.

Original languageEnglish (US)
Pages (from-to)161-183
Number of pages23
JournalPsychometrika
Volume81
Issue number1
DOIs
StatePublished - Mar 1 2016

Fingerprint

Multinomial Model
Probit
Probit Model
Benchmarking
Fractional Factorial Design
Markov Chains
Markov Chain Monte Carlo Algorithms
Prior Information
Simulation Analysis
Scanner
Markov processes
Benchmark
Coefficient
Model
Estimate
Demonstrate

All Science Journal Classification (ASJC) codes

  • Psychology(all)
  • Applied Mathematics

Cite this

Fong, Duncan K.H. ; Kim, Sunghoon ; Chen, Zhe ; DeSarbo, Wayne S. / A Bayesian Multinomial Probit MODEL FOR THE ANALYSIS OF PANEL CHOICE DATA. In: Psychometrika. 2016 ; Vol. 81, No. 1. pp. 161-183.
@article{74d7636c8054489e8fe3b826c52b2a8b,
title = "A Bayesian Multinomial Probit MODEL FOR THE ANALYSIS OF PANEL CHOICE DATA",
abstract = "A new Bayesian multinomial probit model is proposed for the analysis of panel choice data. Using a parameter expansion technique, we are able to devise a Markov Chain Monte Carlo algorithm to compute our Bayesian estimates efficiently. We also show that the proposed procedure enables the estimation of individual level coefficients for the single-period multinomial probit model even when the available prior information is vague. We apply our new procedure to consumer purchase data and reanalyze a well-known scanner panel dataset that reveals new substantive insights. In addition, we delineate a number of advantageous features of our proposed procedure over several benchmark models. Finally, through a simulation analysis employing a fractional factorial design, we demonstrate that the results from our proposed model are quite robust with respect to differing factors across various conditions.",
author = "Fong, {Duncan K.H.} and Sunghoon Kim and Zhe Chen and DeSarbo, {Wayne S.}",
year = "2016",
month = "3",
day = "1",
doi = "10.1007/s11336-014-9437-6",
language = "English (US)",
volume = "81",
pages = "161--183",
journal = "Psychometrika",
issn = "0033-3123",
publisher = "Springer New York",
number = "1",

}

A Bayesian Multinomial Probit MODEL FOR THE ANALYSIS OF PANEL CHOICE DATA. / Fong, Duncan K.H.; Kim, Sunghoon; Chen, Zhe; DeSarbo, Wayne S.

In: Psychometrika, Vol. 81, No. 1, 01.03.2016, p. 161-183.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Bayesian Multinomial Probit MODEL FOR THE ANALYSIS OF PANEL CHOICE DATA

AU - Fong, Duncan K.H.

AU - Kim, Sunghoon

AU - Chen, Zhe

AU - DeSarbo, Wayne S.

PY - 2016/3/1

Y1 - 2016/3/1

N2 - A new Bayesian multinomial probit model is proposed for the analysis of panel choice data. Using a parameter expansion technique, we are able to devise a Markov Chain Monte Carlo algorithm to compute our Bayesian estimates efficiently. We also show that the proposed procedure enables the estimation of individual level coefficients for the single-period multinomial probit model even when the available prior information is vague. We apply our new procedure to consumer purchase data and reanalyze a well-known scanner panel dataset that reveals new substantive insights. In addition, we delineate a number of advantageous features of our proposed procedure over several benchmark models. Finally, through a simulation analysis employing a fractional factorial design, we demonstrate that the results from our proposed model are quite robust with respect to differing factors across various conditions.

AB - A new Bayesian multinomial probit model is proposed for the analysis of panel choice data. Using a parameter expansion technique, we are able to devise a Markov Chain Monte Carlo algorithm to compute our Bayesian estimates efficiently. We also show that the proposed procedure enables the estimation of individual level coefficients for the single-period multinomial probit model even when the available prior information is vague. We apply our new procedure to consumer purchase data and reanalyze a well-known scanner panel dataset that reveals new substantive insights. In addition, we delineate a number of advantageous features of our proposed procedure over several benchmark models. Finally, through a simulation analysis employing a fractional factorial design, we demonstrate that the results from our proposed model are quite robust with respect to differing factors across various conditions.

UR - http://www.scopus.com/inward/record.url?scp=84959242709&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959242709&partnerID=8YFLogxK

U2 - 10.1007/s11336-014-9437-6

DO - 10.1007/s11336-014-9437-6

M3 - Article

C2 - 25491165

AN - SCOPUS:84959242709

VL - 81

SP - 161

EP - 183

JO - Psychometrika

JF - Psychometrika

SN - 0033-3123

IS - 1

ER -