|
Activity Number:
|
512
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #306771 |
|
Title:
|
Flexible Random Intercept Models for Binary Outcomes Using Mixtures of Normals
|
|
Author(s):
|
Ming-Wen An*+ and Brian Caffo and Charles Rohde
|
|
Companies:
|
Johns Hopkins University and Johns Hopkins University and Johns Hopkins University
|
|
Address:
|
3016 Saint Paul Street, Apt. 2, Baltimore, MD, 21218,
|
|
Keywords:
|
probit-normal ; logit-normal ; marginalized multilevel models
|
|
Abstract:
|
Random intercept models for binary data are useful for addressing between-subject heterogeneity. Unlike linear models, the non-linearity of link functions used for binary data force a distinction between marginal and conditional random intercept models. This distinction is blurred in probit models with a normally distributed random intercept because the resulting model implies a probit marginal link as well. We explore another family of random intercept models with this "closure" property. In particular, we consider instances when the distributions associated with the conditional and marginal links and the random effect distribution are mixtures of normals. We relate this flexible family of models to others in the literature and show the associated computational benefits. A diverse series of examples illustrates the wide applicability of the approach.
|
