www.cmm.bristol.ac.uk/forum

Hi Sian, I think you are a little confused here. The DIC diagnostic is an information criterion that can be used to compare models in a similar way to the AIC and BIC in a non-Bayesian/MCMC setting. What it is not is equivalent to 'a two-tailed test at 5% significance level'. The closest thing in th...

Hi Jelle, Yes this sounds appropriate for the analysis you intend to do. I would firstly look at each of the 2 ratings separately and fit a straightforward single response cross-classified model as in the chapter in my MCMC book. To fit a MV model you will need to follow the example in the MV Normal...

This shouldn't be too bad in MCMC. Basically set up your model as in the MCMC manual chapter for multinomial then add on the extra levels as you would anyway. Ignore any error messages from IGLS. Switch to MCMC and tick the cross classified box as in the cross classified chapter in the MCMC manual. ...

Hi Thomasm, If you wish to create imputed datasets to fill in what will ultimately be X variables in your model to fit you need to have them as response variables in the imputation model you first fit. Say for example with the hungary dataset your final model is going to be a regression of biol_core...

Hi Dave, I took a look at your data and discovered that the IGLS algorithm is estimating a non-positive definite variance matrix at level 2. Apologies for the technical jargon but basically if you were to look at your estimates for the between TherID variance matrix in the Estimate Tables window and...

Hi Michael,
Rather impressively you are the first person to spot this mistake - yes the elements given give 90% not 95% confidence intervals. If/When I get around to updating the manual next I'll try and amend this.
Thanks,
Bill.

Hi Karen,
I could take a look if you were happy to send on the worksheet. Note i and j might refer to the areas in the order the software receives them and not the labels they have.
Bill.

Hi NtD71, I suspect what happened is you had a column called 'P' in the data you imported and the macros for multinomial models name certain columns with single letter names including 'P'. Thus when the macro tries to name something as 'P' MLwiN complains that 'P' already exists. The only solution i...

Hi Katy, With MCMC estimation you need to constrain some parameters to make the factor analysis identifiable as the method is simulation-based. The constraining of a loading to 1 is one possible scenario - you will then get loadings of other variables relative to this loading so you will have some m...