This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: PubH 8452 Fall 2011 Homework #3 Due 23 November 2011 1. For multivariate binary data there are several possible parameterizations of the model that have been proposed for regression analysis of the correlated outcomes. All models for multivariate categorical data are special cases of the canonical loglinear model. The difference between models is the way that the joint probabilities are structured in terms of the parameters of interest. (a) Loglinear model (DHLZ 8.2.1) Consider a data set with four binary measurements taken on each of the subject. Define the saturated loglinear model. Give an interpretation for the parameters θ i 1 ,j and θ i 2 ,jk for a given j = 1 , 2 , 3 , 4 and k 6 = j (where i indexes the subject). (b) What is E( Y ij ) in terms of the loglinear model parameters? (c) Bahadur model DHLZ 8.2.2 presents a “fully marginal” model for binary data known as the Bahadur model. For the data above write down the likelihood function based on the Bahadur model with threeway and fourway correlations assumed to be zero. Given an interpretation for the parameters μ ij and ρ ijk for a given j and k 6 = j . (d) Consider a single pair of binary response variables ( Y ij ,Y ik ). Derive the correlation ρ ijk = Cor( Y ij ,Y ik ) such that it is a function of E( Y ij  Y ik = 1) E( Y ij  Y ik = 0). What’s ρ ijk when E( Y ij ) = E( Y ik ) Give a simple interpretation for ρ ijk in this case. (e) Consider an exchangeable correlation matrix for binary Y ij ,j = 1 , ··· ,n i with a single scalar predictor X ij = X i . Please derive the score function of GEE and discuss its relationship to quasilikelihood using a betabinomail variance model for (Σ j Y ij ,n i ). (Note: see DHLZ pages 6061 for R 1 [the notation for R in DHLZ is V ])....
View
Full
Document
This note was uploaded on 11/21/2011 for the course PUBH 8452 taught by Professor Xianghualuo during the Fall '11 term at Minnesota.
 Fall '11
 XianghuaLuo

Click to edit the document details