JSM 2004 - Toronto

Abstract #301419

This is the preliminary program for the 2004 Joint Statistical Meetings in Toronto, Canada. 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, 2004); 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.

Back to main JSM 2004 Program page
Activity Number: 199
Type: Contributed
Date/Time: Tuesday, August 10, 2004 : 8:30 AM to 10:20 AM
Sponsor: Social Statistics Section
Abstract - #301419
Title: Mixed Effect Models and Quasi-F: For Testing Several Conditions in Multiple Random Effect ANOVA Design
Author(s): Olivier Renaud*+ and Paolo Ghisletta
Companies: University of Geneva and University of Geneva
Address: 40, Bd du Pont d'Arve, Geneva 4, International, 1211, Switzerland
Keywords: by item/ by subject ; language-as-fixed-effect fallacy ; hierarchical model
Abstract:

Exact tests for ANOVA exist merely for designs that are limited to one random effect, while multiple fixed effects are allowed. In the social sciences, the random factor is usually the subject. This is especially true in psychology, where these designs are called "repeated measure ANOVA." However, very frequently a second random factor is present, crossed with subject (items presented, such as images, words, etc.). The substantive interest mainly resides in the significance of the fix effect(s) and possible additional covariates of this complex design, after controlling for the random effects. We will review different methods for this problem. In particular, the so-called "by subject/by item" (or F1/F2) method, which is overwhelmingly used in psycholinguistics. The quasi-F method and the mixed effect models methodology will be discussed and compared. The actual level and power of the tests for the different approaches are quite different. Additional discussed issues concern handling of missing values, availability of contrast computations and possibility to have covariates that are linked to one of the random effects.

  • 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 2004 program

JSM 2004 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 2004