Optimal experimental designs for conjoint analysis: Estimation of utility functions

Jose M. Vidal-Sanz, Mercedes Esteban-Bravo, Agata Leszkiewicz

    Research output: Contribution to conferencePaperAcademicpeer-review

    6 Downloads (Pure)

    Abstract

    In conjoint analysis consumers utility functions over multiattributed stimuli are estimated using experimental data. The quality of these estimations heavily depends on the alternatives presented in the experiment. An efficient selection of the experiment design matrix allows more information to be elicited about consumer preferences from a small number of questions, thus reducing experimental cost and respondent’s fatigue. Kiefer’s methodology considers approximate optimal design selecting the same combination of stimuli more than once. In the context of conjoint analysis, replications do not make sense for individual respondents. We present a general approach to compute optimal designs for conjoint experiments in a variety of scenarios and methodologies: continuous, discrete and mixed attributes types, customer panels with random effects, and quantile regression models. We do not compute good designs, but the best ones according to the size (determinant or trace) of the information matrix of the associated estimators without repeating profiles as in Kiefer’s methodology. We use efficient optimization algorithms to achieve our goal.
    Original languageEnglish
    Publication statusPublished - 2012
    Event5th International Conference of the ERCIM Working Group on Computing & Statistics 2012 - Conference Center “Ciudad de Oviedo", Oviedo
    Duration: 1 Dec 20123 Dec 2012
    Conference number: 5

    Conference

    Conference5th International Conference of the ERCIM Working Group on Computing & Statistics 2012
    CityOviedo
    Period1/12/123/12/12

    Fingerprint

    Conjoint analysis
    Methodology
    Utility function
    Experimental design
    Experiment
    Estimator
    Scenarios
    Random effects
    Regression model
    Consumer preferences
    Fatigue
    Consumer utility
    Replication
    Quantile regression
    Experiment design
    Costs

    Cite this

    Vidal-Sanz, J. M., Esteban-Bravo, M., & Leszkiewicz, A. (2012). Optimal experimental designs for conjoint analysis: Estimation of utility functions. Paper presented at 5th International Conference of the ERCIM Working Group on Computing & Statistics 2012, Oviedo, .
    Vidal-Sanz, Jose M. ; Esteban-Bravo, Mercedes ; Leszkiewicz, Agata . / Optimal experimental designs for conjoint analysis : Estimation of utility functions. Paper presented at 5th International Conference of the ERCIM Working Group on Computing & Statistics 2012, Oviedo, .
    @conference{b5dda447e2564bee83df17a8184c8f53,
    title = "Optimal experimental designs for conjoint analysis: Estimation of utility functions",
    abstract = "In conjoint analysis consumers utility functions over multiattributed stimuli are estimated using experimental data. The quality of these estimations heavily depends on the alternatives presented in the experiment. An efficient selection of the experiment design matrix allows more information to be elicited about consumer preferences from a small number of questions, thus reducing experimental cost and respondent’s fatigue. Kiefer’s methodology considers approximate optimal design selecting the same combination of stimuli more than once. In the context of conjoint analysis, replications do not make sense for individual respondents. We present a general approach to compute optimal designs for conjoint experiments in a variety of scenarios and methodologies: continuous, discrete and mixed attributes types, customer panels with random effects, and quantile regression models. We do not compute good designs, but the best ones according to the size (determinant or trace) of the information matrix of the associated estimators without repeating profiles as in Kiefer’s methodology. We use efficient optimization algorithms to achieve our goal.",
    author = "Vidal-Sanz, {Jose M.} and Mercedes Esteban-Bravo and Agata Leszkiewicz",
    year = "2012",
    language = "English",
    note = "5th International Conference of the ERCIM Working Group on Computing & Statistics 2012 ; Conference date: 01-12-2012 Through 03-12-2012",

    }

    Vidal-Sanz, JM, Esteban-Bravo, M & Leszkiewicz, A 2012, 'Optimal experimental designs for conjoint analysis: Estimation of utility functions' Paper presented at 5th International Conference of the ERCIM Working Group on Computing & Statistics 2012, Oviedo, 1/12/12 - 3/12/12, .

    Optimal experimental designs for conjoint analysis : Estimation of utility functions. / Vidal-Sanz, Jose M.; Esteban-Bravo, Mercedes; Leszkiewicz, Agata .

    2012. Paper presented at 5th International Conference of the ERCIM Working Group on Computing & Statistics 2012, Oviedo, .

    Research output: Contribution to conferencePaperAcademicpeer-review

    TY - CONF

    T1 - Optimal experimental designs for conjoint analysis

    T2 - Estimation of utility functions

    AU - Vidal-Sanz, Jose M.

    AU - Esteban-Bravo, Mercedes

    AU - Leszkiewicz, Agata

    PY - 2012

    Y1 - 2012

    N2 - In conjoint analysis consumers utility functions over multiattributed stimuli are estimated using experimental data. The quality of these estimations heavily depends on the alternatives presented in the experiment. An efficient selection of the experiment design matrix allows more information to be elicited about consumer preferences from a small number of questions, thus reducing experimental cost and respondent’s fatigue. Kiefer’s methodology considers approximate optimal design selecting the same combination of stimuli more than once. In the context of conjoint analysis, replications do not make sense for individual respondents. We present a general approach to compute optimal designs for conjoint experiments in a variety of scenarios and methodologies: continuous, discrete and mixed attributes types, customer panels with random effects, and quantile regression models. We do not compute good designs, but the best ones according to the size (determinant or trace) of the information matrix of the associated estimators without repeating profiles as in Kiefer’s methodology. We use efficient optimization algorithms to achieve our goal.

    AB - In conjoint analysis consumers utility functions over multiattributed stimuli are estimated using experimental data. The quality of these estimations heavily depends on the alternatives presented in the experiment. An efficient selection of the experiment design matrix allows more information to be elicited about consumer preferences from a small number of questions, thus reducing experimental cost and respondent’s fatigue. Kiefer’s methodology considers approximate optimal design selecting the same combination of stimuli more than once. In the context of conjoint analysis, replications do not make sense for individual respondents. We present a general approach to compute optimal designs for conjoint experiments in a variety of scenarios and methodologies: continuous, discrete and mixed attributes types, customer panels with random effects, and quantile regression models. We do not compute good designs, but the best ones according to the size (determinant or trace) of the information matrix of the associated estimators without repeating profiles as in Kiefer’s methodology. We use efficient optimization algorithms to achieve our goal.

    M3 - Paper

    ER -

    Vidal-Sanz JM, Esteban-Bravo M, Leszkiewicz A. Optimal experimental designs for conjoint analysis: Estimation of utility functions. 2012. Paper presented at 5th International Conference of the ERCIM Working Group on Computing & Statistics 2012, Oviedo, .