JSM 2005 - Toronto

Abstract #304769

This is the preliminary program for the 2005 Joint Statistical Meetings in Minneapolis, Minnesota. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 7-10, 2005); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.



The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

The Program has labeled the meeting rooms with "letters" preceding the name of the room, designating in which facility the room is located:

Minneapolis Convention Center = “MCC” Hilton Minneapolis Hotel = “H” Hyatt Regency Minneapolis = “HY”

Back to main JSM 2005 Program page
Legend: = Applied Session, = Theme Session, = Presenter
Activity Number: 35
Type: Contributed
Date/Time: Sunday, August 7, 2005 : 2:00 PM to 3:50 PM
Sponsor: General Methodology
Abstract - #304769
Title: Relative Importance and Proportional Marginal Variance Decomposition (PMVD)
Author(s): Barry Feldman*+
Companies: Prism Analytics/DePaul University
Address: 57 W Jackson Blvd. #724, Chicago, IL, 60614, United States
Keywords: Variance decomposition ; Cooperative game ; Statistical significance ; Relative importance ; Proportional value ; Risk analysis
Abstract:

I first propose four intuitive admissibility criteria for measures of the relative importance of independent variables in statistical models. I then consider the problem of identifying a probability distribution over the orderings of independent variables that indicates the likelihood that any ordering is correctly ordered by increasing relative importance. Two technical axioms and the axiom of equal proportional effect identify a unique distribution. The expected contribution to model performance of any variable according to this distribution is the Proportional Marginal Decomposition of the performance measure. When variance is the performance measure, the method effects a variance decomposition. This decomposition is shown equal to the proportional value (Feldman 1999 and 2002, Ortmann 2000) of the cooperative game defined by the marginal variance contributions of all sets of variables. PMVD is shown to be an admissible measure of relative performance. The averaging method of variance decomposition (Lindemen, Merenda, and Gold 1980) and naïve variance decomposition (Pratt 1987) are shown to be inadmissible. Examples and bootstrap confidence intervals are presented.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2005 program

JSM 2005 For information, contact jsm@amstat.org or phone (888) 231-3473. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2005