Preliminary Program subject to change
New Methods for Analyzing and Modeling Complex Data in Interdisciplinary Research
Structure	Measurement
Description	Multi-level and complex data in the field are usually analyzed by conventional statistical methods, which may violate important assumptions, ignore internal associations, and lose rich information within the data; therefore, may generate misleading results and lead to incorrect conclusions. The purpose of this symposium is to introduce four new analytic and modeling methods for multi-level and complex data: Generalized estimating equation, structural mean modeling, social network analysis, and agent-based modeling. After providing an overview of these four methods, their applications in physical activity and health behavior research will be described in detail through several examples and computer software/program demonstrations.
Presider:	Kin-Kit Li, Oregon State University, Corvallis, OR
Speakers:	Yong Gao, University of Illinois-Urbana-Champaign, Urbana, IL Haichun Sun, University Of Maryland, Greenbelt, MD Youngsik Park, University of Illinois-Urbana-Champaign, Urbana, IL Miyoung Lee, University of Illinois at Urbana-Champaign, Urbana, IL
Potential Application of Generalized Estimating Equation Method in Physical Activity Research: A Tutorial Yong Gao, University of Illinois-Urbana-Champaign, Urbana, IL and Weimo Zhu, University of Illinois at Urbana-Champaign, Urbana, IL Testing Latent Means Using Structural Mean Modeling Haichun Sun1, Ang Chen2 and Catherine D. Ennis2, (1)University Of Maryland, Greenbelt, MD, (2)University of Maryland, College Park, MD Social Network Analysis for Studying Complex Social Network Phenomena Youngsik Park, University of Illinois-Urbana-Champaign, Urbana, IL and Weimo Zhu, University of Illinois at Urbana-Champaign, Urbana, IL Agent-Based Modeling of Environmental Impact on Physical Activities Weimo Zhu and Miyoung Lee, University of Illinois at Urbana-Champaign, Urbana, IL

Research Consortium