www.cmm.bristol.ac.uk/forum

For cross-classified models the ordering of the residuals should match the sorted ID variable for the chosen classification. See the topic Residuals - matching for an MLwiN command to generate a new column that matches this.

Unfortunately we do not offer an automatic way to calculate these predictions. I think that your options are :


Remove the nopause option from the command and the once the model is estimated use Model>Customised predictions within MLwiN. You can then copy/paste these results back into Stata ...

There is some discussion of this in the MLPowSim manual. Multiple category predictors for MLwiN are covered in section 5.2, and for R in 5.4. You may in particular want to look at section 5.4.2 for an explanation.

The MISR command is used in conjunction with post-estimation calculation of the residuals and are not normally used in model estimation except with PQL1 or PQL2, or in MCMC where they are used for starting values. The runmlwin equivalent of this setting for post-estimation residuals is the norecode ...

If you run the model without the nopause option then MLwiN will pause at various points to allow you to inspect the model/data. You can can then resume it by clicking "Resume macro" on the toolbar. Once you have got through all the points that it pauses the results will be sent to Stata. When you ...

As this section of the forum is for questions related to the Stat-JR software package are you able to provide any more detail on how you are using it for this analysis, e.g. which templates and underlying statistics package you are using/planning to use?

You appear to have omitted your code, however if you want to specify CAR at more than one level you have to specify the carcentre option.

You could try a Wald test, an example of which is given in the MLwiN manual replication examples for runmlwin:

Chapter 9: Logistic Models for Binary and Binomial Responses

A more detailed explanation for these examples can be found in the corresponding chapter of the MLwiN User Guide

Unfortunately I have not had much success with running the MLwiN trial under Wine on MacOS versions later than 10.14, as later versions removed support for 32-bit applications and then moved from the Intel to ARM platform. This might change in the future as Wine is developed and experimental support ...

It's hard to say what exactly is causing this error without seeing the data. It could be something like there being colinear variables in the model, or you might have categorical variables where no data is present for a particular category value.

In order to work out what might be going on I would ...