|
Activity Number:
|
64
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #303878 |
|
Title:
|
Longitudinal Bayesian Analysis of Self-Reported Noisy Count Data
|
|
Author(s):
|
Jihey Lee*+ and Robert E. Weiss
|
|
Companies:
|
University of California, Los Angeles and University of California, Los Angeles
|
|
Address:
|
Department of Biostatistics, School of Public Health, Los Angeles, CA, 90095,
|
|
Keywords:
|
rounding error ; longitudinal count data ; Bayesian inference
|
|
Abstract:
|
In self-reported data, reported counts of events are not accurate due to many reasons: subjects may not recall well; they may round the counts; they may over- or under- report. These types of recording errors cause overdispersion and make a Poisson model a poor fit. Ignoring recording error and using reported values as the direct outcome variable in the model can lead to biased parameter estimates and conclusions that are not supported by the data. We describe a Bayesian model for analyzing longitudinal count data, where bias resulting from recording error is corrected by a stochastic process model and a random effects model is fitted to the underlying 'true' values. The proposed methodology will be applied to model the number of sex partners in last 3 months of HIV-infected youths whose HIV transmission behaviors were examined over 15 months after receiving a preventive intervention.
|
