|
Activity Number:
|
459
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #306503 |
|
Title:
|
Are Higher Levels of Multilevel (Hierarchical) Models Necessary? Application to High-Risk Sexual Behavior Data
|
|
Author(s):
|
DeMarc Hickson*+ and Lance Waller and Lillian Lin
|
|
Companies:
|
Emory University/Centers for Disease Control and Prevention and Emory University and Centers for Disease Control and Prevention
|
|
Address:
|
1600 Clifton Road, Atlanta, GA, 30329,
|
|
Keywords:
|
multilevel models ; hierarchical data ; deviance information criterion
|
|
Abstract:
|
Community Intervention Trial for Youth (CITY) project personnel conducted venue-based interviews of young men who have sex with men (MSM) for four consecutive summers in 12 communities throughout the United States as part of the evaluation of a community-level intervention field trial. This venue-based sampling introduced four levels to the data structure: individual, venue, community, and time. Multilevel (hierarchical) models represent datasets with two or more levels of clustering and allow investigators to examine cross-level effects and include level- or cluster-specific information (e.g., individual and/or macro group). The purpose of this research is to determine if higher levels of (hierarchical) data structures are necessary. We perform simulation studies to the degree to which the deviance information criterion (DIC) will assess the need for higher-data clusters.
|
