|
Activity Number:
|
549
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Teaching Statistics in the Health Sciences
|
| Abstract - #309848 |
|
Title:
|
A Graphical Method for Assessing Mediating Variables Using Brunswik's Lens Model
|
|
Author(s):
|
Matthew P. Spackman*+ and Dawson Hedges and Bruce Brown
|
|
Companies:
|
Brigham Young University and Brigham Young University and Brigham Young University
|
|
Address:
|
1001 SWKT, Provo, UT, 84602,
|
|
Keywords:
|
multivariate graphics ; lens model ; Brunswik
|
|
Abstract:
|
The Brunswikian lens model (1956) is an effective tool for assessing the extent to which accuracy in a classification task is mediated by a set of measurements. The multivariate structure of the mediating variables in the lens model is substantially clarified by the use of multivariate graphics adapted to lens model analysis. The power of this multivariate graphical approach to the lens model is demonstrated with examples from research on the accuracy with which humans can identify emotions from voice samples, with acoustic variables as the mediators. Graphical lens model analysis is further demonstrated with a second example, using volumetric MRI data from various brain locations as mediators of the clinical accuracy of identification of post traumatic stress syndrome.
|
