|
Activity Number:
|
514
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #310004 |
|
Title:
|
Joint Modeling of Longitudinal Binary Outcome and Time to Drop-Out: An Application Using Complex Survey Data
|
|
Author(s):
|
Punam Pahwa*+ and Chandima Karunanayake
|
|
Companies:
|
University of Saskatchewan and University of Saskatchewan
|
|
Address:
|
103 Hospital Drive, Saskatoon, SK, S7N OW8, Canada
|
|
Keywords:
|
Generalized estimation equations ; Bootstrap ; Dropout ; Complex Survey ; Joint Model
|
|
Abstract:
|
Statistics Canada National Population Health Survey (NPHS) data were used to trace longitudinal changes in the mental health of immigrant women. In the NPHS a member from each household was randomly chosen to participate in this longitudinal survey. We explored the moderating affects of different risk factors on immigrant women's mental distress, which was categorized as a dichotomous variable: no/low distress vs. moderate/high distress. Regression coefficients were estimated by using the weighted generalized estimating equations approach which accounts for within-subject correlation and variance estimation was conducted by taking into account the complexities of multi-stage design using bootstrap weights computed by Statistics Canada. An alternative approach based on the joint model for a longitudinal dichotomous outcome and the time to dropout is under process.
|
