|
Activity Number:
|
45
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 11, 2002 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Social Statistics Section*
|
| Abstract - #301347 |
|
Title:
|
Latent Covariate Analysis Using Factor Score Estimate
|
|
Author(s):
|
Kari Azevedo*+ and Yasuo Amemiya+
|
|
Affiliation(s):
|
Iowa State University and IBM T. J. Watson Research Center
|
|
Address:
|
2625 North Loop Suite 500, Ames, Iowa, 50010-8296, USA 33-221 Route 134, Yorktown Heights, New York, 10598,
|
|
Keywords:
|
analysis of covariance ; structural equation modeling ; nonlinear relation ; intervention study
|
|
Abstract:
|
A general approach is proposed for assessing the effectiveness of comparative treatment outcomes in social and behavioral studies. In such studies, designed as experimental rather than observational, modeling of relations using conventional structural equation techniques may not be directly relevant for comparing different programs. Since the latent intervention treatment effect is often ambiguous, one must consider ways to improve sensitivity of the intervention assessment. The use of covariates can improve the power of comparison tests and reduce the width of a confidence interval for a treatment difference. Most intervention studies contain observed measurements unaffected by treatments that are related to the intervention-targeted response which can be used for this purpose. Also, relations between the response and underlying covariates may not always be linear. We propose an overall approach to the latent covariate analysis that uses factor score estimates. We develop a coherent analysis of covariance procedure that can examine various nonlinear covariate effects and provides an efficient and proper assessment of the intervention comparison.
|
- 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 2002 program |
