www.cmm.bristol.ac.uk/forum

Simply add an extra level i.e. to fit a 3 level multivariate model in MLwiN one simply asks for 4 levels with the additional lowest level being used to represent the responses within an individual. Hope that helps.
Bill.

Hi GijsHuitsing, But the issue here is there is a 1-1 mapping of outcomes to classes so probably no need to use multiple membership unless you are going to included some sort of cumulative effects for earlier class-rooms. One uses a multiple membership model if say you only had response at wave 3 bu...

Hi GijsHuitsing, I am unclear whether you have multiple membership or simply cross-classification? If your structure is that each measurement is nested within 1 student and 1 class but the classes change for measures then this is simply a cross-classification and you simply tick the cross-classified...

Dear Xinyuzou, There are many reasons why mixing might be poor and different methods e.g. orthogonal parameterisation and hierarchical centering will fix different issues. Other things to consider are centering predictor variables. It may end up that you have to run for longer. The paper by Browne e...

Hi Shane,
My immediate advice is to contact Joop directly as he may have a macro to do this in MLwiN which he has been known to use. Without reading Joop's book which I don't have time to as we speak I can't help you any further.
Apologies,
Bill.

HI Shane, It looks like in MQL1 that although a household variance is picked up it has a very large standard error which might suggest uncertainty that there are really significant household effects. I would suggest checking that your hierarchy is correct i.e. that the data is sorted correctly. I wo...

Dear Edward, Apologies for the delay in replying - I think this might be a case of 'all models are wrong but some models are interesting' as George Box is quoted. Your normalised scores plot does indeed show something of an S shape which I am guessing is indicating the threshold at the top end. Of c...

Hi Edward, MLwiN does not have specific functionality for handling censored data. In fact the exam datasets that are often used in educational examples have similar issues of censoring - where marks are assumed normally distributed but in reality are constrained to lie between 0 marks and full marks...

Hi Frank, The level 2 variation in a binary response model is on a different scale (the logistic scale) from the response (the probability scale) and it is well known that when one adds predictors to a model this can have the effect of rescaling parameters so that the underlying binomial assumption ...

Hi Sasso, Post-hoc power calculations are generally frowned upon in the stats community - see for example http://research-repository.st-andrews.ac.uk/bitstream/10023/679/5/Thomas-Retrospectivepoweranalysis-postprint.pdf although in practice we often talk about using pilot studies to get estimates fo...