www.cmm.bristol.ac.uk/forum

Hi Ash, I suspect the problem is associated with the fact that an N level multivariate response model is formulated as a N+1 level model in MLwiN where the lowest level is a pseudo level. (There are no random effects at this level, it exists solely to define the structure of the multivariate respons...

That's great. I'm glad it worked.

Thank you for putting all the graphs and code in. It makes this post a useful resource for other users

Best wishes

George

Hi Ewan, This is also an excellent question and follows naturally from your previous question. The answer is similar to before: After running runmlwin , you can generate all the predictions AND THEIR SAMPLING ERRORS (AND THEREFORE CONFIDENCE INTERVALS) that you would normally make using the "Pr...

Hi Ewan, This is a very good question. After running runmlwin , you can generate all the predictions that you would normally make using the "Predictions" or "Customised Predictions" windows in MLwiN. However, we have not programmed up a runmlwin specific prediction post-estimatio...

Hi Ash, What has happened is that the 2-by-2 student-level variance-covariance matrix has gone non-positive-definite on one of the iterations. What MLwiN does to solve this problem is to reset the relevant row of the variance-covariance matrix to zero and then to carry on iterating until convergence...

Hi Paula, Part of your query asked how you can enter a categorical variable into a runmlwin model as a series of dummy variables without having to manually generate these dummy variables. One way to do this is to use the xi: prefix and the i. operator. For example: . use http://www.bristol.ac.uk/cmm...

Hi Laia, Thank you. You spotted a bug. We were not calculating the lower and upper bounds of the 95% confidence intervals for odds ratios in the correct way. We were calculating them as plus or minus 1.96 times the standard error of the odds ratio. This is incorrect as it assumes the sampling distri...

Dear J, This is a good question. No we don't plan to extend Stata's postestimation predict command to work with runmlwin . This would involve working out and implementing the prediction formula for a very wide range of multilevel models. The aim of runmlwin is to make MLwiN functionality accessible ...

Hi Corrie, Thanks for spotting this. Yes runmlwin is limited in how many explanatory variables you can put into the model. You can see this by running the following code: clear set obs 1000 gen id = _n gen y = rnormal() gen cons = 1 forvalues i = 1/200 { gen x`i' = rnormal() } quietly runmlwin y x1-...

This bug is now fixed.

To get the latest version of runmlwin, type the following in a net-aware version of Stata:

. adoupdate runmlwin

Best wishes

George