The model did not run properly in MLwiN. You most likely clicked the 'Abort Macro' button in MLwiN, rather than the 'Res
-
- Posts: 1354
- Joined: Mon Oct 19, 2009 10:34 am
Re: The model did not run properly in MLwiN. You most likely clicked the 'Abort Macro' button in MLwiN, rather than the
When fitting discrete response models using (R)IGLS in MLwiN the estimates are fitted using a quarsi-likelihood method. As this is an approximate method a Log-Likelihood value is not calculated (see http://www.bristol.ac.uk/cmm/software/s ... equawindow). In general we recommend fitting at least your final versions of these models using MCMC. For further information on fitting discrete response models in MLwiN see http://www.bristol.ac.uk/cmm/software/m ... l#discrete.