mixed effect models in R using glmer

#1
Folks, I'm looking for a very basic introduction to mixed effect models in a health care context. Suggestions welcome.


Thanks
 

hlsmith

Omega Contributor
#2
What are you thinking, panel, cross -sectional, cross -classified data and how would the dependent variable be formatted?
 
#3
What are you thinking, panel, cross -sectional, cross -classified data and how would the dependent variable be formatted?
Here's a look at the data. Response variable is Y, and trying to tease out hospital and surgeon effects.

hospital surgeon patient year procedure sex race ageGrp Y
1 90001 4946208438 105764497 2009 4 1 1 3 0
2 210022 4399865965 105764795 2009 5 1 1 4 0
3 210019 4731440957 105764824 2009 5 1 1 4 1
4 210022 4142258806 105764833 2009 5 1 2 3 0
5 490059 4309138380 105764850 2009 5 1 1 4 0
6 210051 4926747245 105765209 2009 4 0 1 4 1

ageBin raceBlack raceOther
1 1 0 0
2 1 0 0
3 1 0 0
4 1 1 0
5 1 0 0
6 1 0 0
 

hlsmith

Omega Contributor
#4
I dont know of a book, not a heavy R user, but it looks like you may be investigating a binary outcome (e.g. logistic) and may want to have a 2nd and 3rd level for these cross -sectional data.

Hope that helps get you closer to the right R function (maybe lme4).

PS, if those are real hosp# you may want to strip them out of the post.
 
#6
I dont know of a book, not a heavy R user, but it looks like you may be investigating a binary outcome (e.g. logistic) and may want to have a 2nd and 3rd level for these cross -sectional data.

Hope that helps get you closer to the right R function (maybe lme4).

PS, if those are real hosp# you may want to strip them out of the post.
Hey, thanks. All the data is fake except the hospital numbers. Yes, it's a binary outcome.