I would like to specify more than one random effect in a generalized linear mixed model with glmmPQL (MASS package). Currently I am using the data from Heck, Thomas and Tabata (2012).
So what I would like to achieve is to prove, if the slopes of the level 1-predictors (femal, minor and lowses) vary between the level 2 units.
Is this the right code to do so?
If not – How can I specify multiple random effects simultaneously?
PQL<-glmmPQL(readprof~female+minor+lowses+schcomp+smallsch, random=(~1|schcode/minor/female/lowses), family = binomial(link = logit), data = Heck) Many thanks!
Best Answer
So what I would like to achieve is to prove, if the slopes of the level 1-predictors (femal, minor and lowses) vary between the level 2 units.
I'm new to glmmPQL and mixed-effects models, and I"m not entirely sure what you mean by "level 2 units" here, so I can't check that your model is constructed properly. However, I think I can help you with your question:
How can I specify multiple random effects simultaneously?
According to the glmmPQL documentation, you need to create a matrix of the formulas to input multiple random effects.
PQL<-glmmPQL(readprof~female+minor+lowses+schcomp+smallsch, random=list(schcode = ~1, minor = ~1, female = ~1, lowses = ~1), family = binomial(link = logit), data = Heck) I found the above info at stackoverflow.
Slashes are used for covariates in a model, rather than just allowing you to input the multiple random effects. You can see how they work in the link.
I hope this helps.
Similar Posts:
- Solved – Variance structure with GLMMPQL
- Solved – treat SNP genotype 0 1 2 as continuous value
- Solved – How to test (in Stata) whether the gender distribution of employees to jobs differs across two companies
- Solved – Providing starting values for a Generalized Linear Mixed Model with glmmPQL
- Solved – Adding a variance structure when fitting a gamm with Gamma distribution