|
Activity Number:
|
381
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #308256 |
|
Title:
|
Bayesian Analysis of Repeated Behaviors Using Event-Dependent Frailty Models
|
|
Author(s):
|
Getachew Dagne*+
|
|
Companies:
|
University of South Florida
|
|
Address:
|
13201 Bruce B Down Blvd, Tampa, FL, 33612,
|
|
Keywords:
|
Event dependent ; emotion regulation ; random effects ; Weibull model ; Bayesian inference ; social interaction
|
|
Abstract:
|
In social behavioral studies, we commonly encounter repeated displays of behaviors (events) data. Statistical methods for the analysis of such data use either parametric or semi-nonparametric proportional hazards model, modified to include frailty which summarizes the dependence of recurrent times within a dyad. These random frailty effects typically model heterogeneity among dyads. However, dyad-specific random frailty may not capture event dependence that occurs because of ordering of event occurrences. The occurrence of one behavior (event) may make further occurrence of the behavior more or less likely based on the characteristics of dyads. This paper presents an event-dependent frailty model for recurrent sojourn time data in the Bayesian context. Our approach is illustrated by a dataset relating to families with children who show emotional regulation problems.
|
