688 – Personalized Intervention Based on Health Care Big Data Research
On the Impact of Informative Nonresponse in Logistic Regression
Joanna JJ Wang
University of Technology Sydney
Mark Bartlett
The Sax Institute
Louise Ryan
University of Technology
In this paper, we are interested in nonignorable missing data mechanism where the probability of nonresponse depends on the outcome. We consider a selection model for nonignorable nonresponse in logistic regression. Expressions for the bias in parameter estimates are derived in a simple case. Further, we propose a sensitivity analysis to study changes in parameter estimates under different assumptions. We adopt a Bayesian framework as it offers a flexible approach for incorporatin different missing data mechanisms. Our modelling strategy is illustrated using survey data from the 45 and Up Study.