New Methods for the Analysis of Family and Dyadic Processes
SATURDAY October 14---SUNDAY October 15, 2006
Welcoming reception: FRIDAY, Oct. 13, 5:30-7:30 p.m.
Poster Session and Reception: SATURDAY, Oct. 14, 5:30-7:30 p.m.
The conference brought together experts in psychology, sociology, education, and biostatistics to examine state-of-the art statistical methods that address the special issues that arise in research on processes within families and small groups. Common issues include statistical non-independence, mediational processes in clustered data, endogeneity and selection bias in longitudinal models, analyzing intensive longitudinal diary data, and distinguishing exchangeable and non-exchangeable dyads.
The conference was attended by substantive researchers in social and relationship research and family science interested in learning about and applying these new methodologies. It focused on discussion of recent innovations in an accessible format and also featured a forum for students and postdoctoral researchers to share their current work through a poster session. Complete poster abstracts may be found here.
SELECTED TOPICS and SPEAKERS
* Causal inferences with group-based trajectory methods
Daniel Nagin, Carnegie-Mellon University
* Modeling dyadic and family processes as dynamical systems
Steven Boker, University of Notre Dame
* Extensions of growth modeling to applications with binary and zero-inflated variables
Alan Acock, Oregon State University
* Statistical and methodological issues in the use of diary methods to study dyadic and family processes
Jean-Philippe Laurenceau, Univ of Delaware
* Moderator Effect in the Actor-Partner Interdependence Model
David Kenny, University of Connecticut
* Mediation in multilevel models for family and dyadic data
Niall Bolger, Columbia University
* Structural equation models for exchangeable and distinguishable dyads and small groups
Joseph Olsen, Brigham Young University
* Multiple imputation for incomplete multilevel data with SHRIMP
Recai Yucel, Univ. of Mass Amherst
* Multivariate regression models for analyzing data from multiple informants
Garrett Fitzmaurice, Harvard University
* Comparing group effects in logit and probit models
Scott Long, University of Indiana
* Trajectories of Physiological Measures in Longitudinal Dyads
Sally Powers, University of Massachusetts Amherst
Aline Sayer, Univ. of Mass Amherst
* Applications of item response theory to improve health outcomes assessment
Ronald Hambleton, Univ. of Mass Amherst
Aline Sayer and Sally Powers
The conference was hosted by the Center for Research on Families in the College of Social and Behavioral Sciences at the University of Massachusetts Amherst, and co-sponsored by the Office of the Vice Provost for Research through a Research Leadership in Action grant and by the Science Directorate of the American Psychological Association (APA).