|
Activity Number:
|
390
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #310184 |
|
Title:
|
A New Model for Interaction Effects
|
|
Author(s):
|
David Weakliem*+
|
|
Companies:
|
University of Connecticut
|
|
Address:
|
344 Mansfield Rd, Storrs, CT, 06269-2068,
|
|
Keywords:
|
interaction ; general linear model
|
|
Abstract:
|
Interaction effects are often of interest in the social sciences. When the variables involved have a moderate or large number of categories, imposing some structure on the interactions makes results easier to interpret and allows more powerful tests against the additive model. Moreover, different hypotheses about social processes imply different forms of interaction effects. This paper discusses some structured models for two-way interaction effects in the general linear model. The idea of competition among variables (for example, ethnicity and social class) has been important, but is not adequately represented by any existing model for interactions. I propose a model representing the interaction resulting from competition and apply it to a number of examples. The paper also discusses the relationship of the proposed model to several existing models for square tables.
|
- 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 2007 program |
