MNL63: Appendix O: Bootstrapping Penalty Analysis

    Xiong, Rui
    University of Arkansas, Fayetteville, AR

    Meullenet, Jean-Francois
    University of Arkansas, Fayetteville, AR

    Pages: 4    Published: Jan 2009


    Abstract

    Bootstrapping penalty analysis is a method that allows you to perform statistical testing on the results of a penalty analysis, using a technique called bootstrapping to estimate variances and covariances. Recall that the purpose of penalty analysis is to identify attributes that appear to have a strong impact on Overall Liking, on a product-by-product basis.. However, there are no significance procedures to test the significance of each attribute's effect on Overall Liking. The bootstrap method, is a well-established computer-based Monte Carlo technique for determining an estimate for the standard errors, confidence intervals, biases, and prediction errors.


    Paper ID: MNL11496M

    Committee/Subcommittee: E18.03

    DOI: 10.1520/MNL11496M


    CrossRef ASTM International is a member of CrossRef.

    ISBN10: 0-8031-7010-6
    ISBN13: 978-0-8031-7010-0