Bayesian Statistics and Marketing

Rossi, Peter E.; Allenby, Greg M.; McCulloch, Robert E.

doi:10.1002/0470863692

Cited by 905 publications

(840 citation statements)

References 0 publications

Supporting

Mentioning

817

Contrasting

Unclassified

Order By: Relevance

“…In the statistical literature, this kind of posterior predictive simulation is the standard approach (Rossi et al 2005). We now turn to the specific simulations and the results.…”

Section: Counterfactual Simulations Of Market Demandmentioning

confidence: 99%

“…To pool data across respondents j=1,2,…J while allowing for heterogeneity of preferences, we use the standard hierarchical approach following Lenk et al (1996). Please see Rossi et al (2005) for an overview of hierarchical linear models. A row vector of M characteristics Zj characterizes each respondent, and respondents with similar characteristics tend to have similar preferences following a multivariate regression:…”

mentioning

confidence: 99%

“…To estimate the parameters of our choice model, we follow a standard Bayesian procedure to generate draws from the posterior distribution of all parameters using a Gibbs sampler. Please see Rossi et al (2005) for a detailed description of setting up the Gibbs sampler for a hierarchical linear model. We ran the Gibbs sampler for 50,000 iterations, discarding the first 10,000 as burn-in iterations and using the remaining 40,000 draws to conduct our counterfactual exercises.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Consumer Preferences for Annuity Attributes: Beyond Net Present Value

Shu

Zeithammer

Payne

2016

Journal of Marketing Research

View full text Add to dashboard Cite

Decisions about life annuities are an important part of consumer decumulation of retirement assets, yet are relatively underexplored by marketing researchers studying consumer financial decision-making. In this paper we propose and estimate a model of individual preferences for life annuity attributes using a choice-based stated-preference survey. Annuities are presented in terms of consumer-relevant attributes such as monthly income, yearly adjustments, period certain guarantees, and company financial strength. We find that attributes directly influence preferences beyond their impact on the annuity's expected present value. The strength of the direct influence depends on how annuities are described: when represented only via basic attributes, consumers undervalue inflation protection and preferences are not monotonically increasing in duration of period certain guarantees. When descriptions are enriched with cumulative payment information, consumers no longer undervalue inflation protection, but nonlinear preferences for period certain options remain. We find that among annuities with the same expected payout but different annual increases and period certain guarantees, the proportion of consumers choosing the annuity over self-management can vary by more than a factor of two. With baby boomers now retiring at the rate of almost 10,000 per day, the issue of decumulation of retirement assets is increasingly important to economists, public policy experts, and the financial services industry. It should also be of interest to researchers in marketing because consumers in the market for decumulation products, such as annuities, face a choice problem with large financial stakes, limited learning opportunities, difficult consumption tradeoffs, multiple sources of uncertainty, issues of trust and branding, and long time periods. All of these aspects of the decumulation problem are topics on which marketing research can offer important insights. This paper studies the structure of consumer preferences for life annuities -an important class of decumulation products. We employ a choice-based conjoint analysis to measure consumer preferences and relate them to the underlying financial value of the products. KeywordsAnnuities, as well as many other financial products, provide a unique setting for choice modeling because most annuity attributes have calculable expected present value that can be directly compared to consumers' revealed utilities. Consequently, we are able to see whether an attribute influences demand only through its contribution to the normative net present financial value of the annuity product ("NPV"), or whether attribute values have psychological worth "beyond NPV." We find that a typical consumer choosing from a set of annuities does not merely maximize the expected financial value, but also reacts to several product attributes directlyexpressing preferences beyond the effect of attributes on the financial value. For example, most consumers overvalue medium (10-20 years) levels of period-certa...

show abstract

“…In the statistical literature, this kind of posterior predictive simulation is the standard approach (Rossi et al 2005). We now turn to the specific simulations and the results.…”

Section: Counterfactual Simulations Of Market Demandmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Consumer Preferences for Annuity Attributes: Beyond Net Present Value

Shu

Zeithammer

Payne

2016

Journal of Marketing Research

View full text Add to dashboard Cite

show abstract

“…A stated-choice experiment reveals the influences of the new safety metric, as if it were available to consumers, along with other attributes, and the data are used to estimate a random coefficients multinomial logit choice model [14] using a hierarchical Bayes techique [15], which projects the vehicles consumers will choose based on the attributes of the alternatives. The attributes are country of origin, fuel economy, personal safety, acceleration performance, two-or four-wheel drive, manual or automatic transmission, two or three rows of seats, and purchase price.…”

Section: Estimation Of Consumer Choicementioning

confidence: 99%

A market systems analysis of the U.S. Sport Utility Vehicle market considering frontal crash safety technology and policy

Hoffenson

Frischknecht

Papalambros

2013

Accident Analysis & Prevention

View full text Add to dashboard Cite

“…However, for many important applications the MNL model takes a more general form, where the choice probabilities contain regression coefficients that are not category specific. Examples include discrete choice models in marketing (Rossi, Allenby, and McCulloch, 2005) and the partial credit model, used in large educational assessment programs such as PISA (Fox, 2010). In its most general form, the probability that y i takes the value k is modeled for k = 1, .…”

Section: Introductionmentioning

confidence: 99%

Bayesian Inference in the Multinomial Logit Model

Frühwirth‐Schnatter

Frühwirth

2016

AJS

View full text Add to dashboard Cite

The multinomial logit model (MNL) possesses a latent variable representation in terms of random variables following a multivariate logistic distribution. Based on multivariate finite mixture approximations of the multivariate logistic distribution, various data-augmented Metropolis-Hastings algorithms are developed for a Bayesian inference of the MNL model.Zusammenfassung: Das multinomiale logistische (MNL) Regressionsmodell besitzt eine latente Variablendarstellung, die einen zufälligen Fehlerterm beinhaltet, der einer multivariaten logistischen Verteilung folgt. Aufbauend auf einer finiten Mischungsapproximation der multivariaten logistischen Verteilung werden mehrere Metropolis-Hastings-Verfahren für eine Bayes-Analyse im MNL Regressionsmodell entwickelt.

show abstract

Bayesian Statistics and Marketing

Cited by 905 publications

References 0 publications

Consumer Preferences for Annuity Attributes: Beyond Net Present Value

Consumer Preferences for Annuity Attributes: Beyond Net Present Value

A market systems analysis of the U.S. Sport Utility Vehicle market considering frontal crash safety technology and policy

Bayesian Inference in the Multinomial Logit Model

Contact Info

Product

Resources

About