www.cmm.bristol.ac.uk/forum

Is it the first or second of your two model fits that triggers the error? If it is the first have you checked whether switching from pql2 to mql1 allows it to fit?

Thanks for confirming that you have found the reason for the problem that you were encountering. When we send the data to MLwiN from R we do so via a Stata .dta file, which I don't believe has a representation for a character missing value. I'd have to look into exactly what is going on but I would ...

Are you able to provide any further details regarding the models that you are trying to fit? Is the goal to fit them within MCMCglmm?

MLwiN does not estimate a level-1 variance for binomial models, instead constraining it to a fixed value. For more details see:

Apportioning Variation in Multilevel Models

and

Partitioning variation across levels

Unfortunately, other than in conjunction with multiple-membership models, MLwiN does not support weights when using MCMC estimation.

I asked George about this and he recommended the following method of replicating the model in MLwiN: // create variable to represent the intercept generate cons = 1 // combine weight and sbp responses into a single column capture drop y replace weight = 0 if weight == . replace sbp = 0 if sbp ==. ge...

We don't have a lot of information on this outside the advanced guide. What kind of extensions are you hoping to add?

MLwiN allows the worksheet columns to have different lengths, however any that are used in the model (including level identifiers, etc) must contain the same number of rows. The error message suggests that this is not the case for your model so I would suggest looking the the Data Manipulation>Names...

I checked this with George and the behaviour is expected and is covered in Snijders and Bosker . George gives the following explanation for the simplest multilevel case (two-level variance components model): The intercept is interpreted as the population average of the cluster-specific population me...

The loading procedure should be the same for both models, I think that you just don't see the loading message on the four level model because the data finishes loading before the message is displayed. The question is therefore why the data takes so much longer to load for the five level model. My su...