The Indonesian children's health study is a longitudinal study.

The data can be downloaded from the website of Dr. Heagerty, see the link below.

http://faculty.washington.edu/heagerty/Books/AnalysisLongitudinal/datasets.html

The objective of this assignment is to estimate the increase in risk of respiratory infection for

children who are vitamin A deficient while controlling for other factors.

1.

Fit two marginal models using the function (geeglm) from the package "geepack". Use

corstr = ar1 for model 1, corstr = unstructured for model 2. Interpret the risk effects

from R output and briefly describe the difference of the results from the two models.

2.

Fit two generalized linear mixed effects models using glmer, with model 1 including a

random intercept, model 2 including both a random intercept and a random slope of

"time". Test if the random slope term is necessary.

3.

Briefly describe the difference between the results from marginal models and random

effects models.