Which survey weights should I use?
Title 

Choosing survey weights for gllamm 
Author 
Minjeong Jeon, University of California, Berkeley 
Date 
July 2012 
The answer depends on what kind of sampling weights you have in your data set.
Suppose you have a dataset consisting of two levels.
We can think of the total weights (w_{ij}) as product of leve12 weights (w_{j}) and level1 weights (w_{ij}):
w_{ij} = w_{j} × w_{ij}
where w_{j} is the inverse of the probability that the level2 unit
(or primary sampling unit) was selected and w_{ij} is the inverse
of the conditional probability that the level1 unit was selected given
that the level2 unit that it belongs to was selected.
Use the level1 and level2 weights if you have both of them.
If only level2 weights are available (and the level1 units were sampled
from level2 units with equal probabilities), use the level2 weights only.
Similarly, if you have level1 weights only and the level2 units
were sampled with equal probabilities, use level1 weights only.
If you have level1 weights, do not use them without proper rescaling of the weights.
Scaling weights is still an ongoing research area; see for example RabeHesketh and Skrondal (2006).
A problem arises when you have only total weights.
If you have only total weights, do not use the total weights for pweight() option.
Remember that pweight() allows weights for individual levels only.
When weights at individual levels are not available, one alternative way is
to utilize design variables. By using the design variables as covariates in your regression model, you can still
obtain consistent estimates. Another alternative is not to use multilevel modeling but
to use ordinary regression models with the svy prefix command.
Examples and documentation
References
