Search found 26 matches

by joneskel
Thu Mar 30, 2017 8:14 pm
Forum: MLwiN user forum
Topic: Predicted probabilities for higher-level units in multinomial models
Replies: 1
Views: 3654

Re: Predicted probabilities for higher-level units in multinomial models

Several aspects to this 1) you can use coverage intervals in customised predictions to give between higher level unit overall differences on the probability scale but this will not give specific hospitals 2 You can use residuals at the higher level to get estimates of the specific hospital differenc...
by joneskel
Tue Mar 14, 2017 9:06 am
Forum: MLwiN user forum
Topic: Simple slopes analyses
Replies: 2
Views: 4072

Re: Simple slopes analyses

I am a bit confused by your questions. and I would not regard 3 way cross - level interactions as simple! The variances are stored iand there are many ways of getting at them. All MlwiN our resources are set out here http://www.bristol.ac.uk/cmm/software/mlwin/mlwin-resources.html and I suggest that...
by joneskel
Fri Mar 25, 2016 9:19 am
Forum: MLwiN user forum
Topic: Random coefficient models - testing for significant reduction
Replies: 1
Views: 3802

Re: Random coefficient models - testing for significant reduction

1 Use a Wald test of whether a parameter or a set of parameters are zero – a univariate Wald and a multivariate Wald (for multiple simultaneous testing) are available under Interval and Tests; in the main you need to be careful with random effects estimates but the fixed ones are generally fine; 2 U...
by joneskel
Fri Mar 25, 2016 9:10 am
Forum: MLwiN user forum
Topic: IGLS/RIGLS numeric warning
Replies: 1
Views: 4409

Re: IGLS/RIGLS numeric warning

One method of estimation in MLWIN is a likelihood based one and there are 2 versions. Iterative Generalised Least Squares estimates the random part under the assumption that fixed part estimates are known. Restrictive IGLS estimates the random part variances as if the fixed part means have been esti...
by joneskel
Tue Jan 19, 2016 2:03 pm
Forum: MLwiN user forum
Topic: Assessing the added value of (logistic) multilevel models
Replies: 2
Views: 4405

Re: Assessing the added value of (logistic) multilevel models

In a discrete outcome model there is no level 1 variance to be estimated as it really depends on the predicted underlying response that is the fixed part. In fitting a separate model model to each country you are doing the equivalent of a fixed effect analysis (see Bell and Jones below) in which in ...
by joneskel
Sat Nov 14, 2015 3:50 pm
Forum: MLwiN user forum
Topic: Out-of-sample predictions: Help & guidance
Replies: 2
Views: 4984

Re: Out-of-sample predictions: Help & guidance

If you want to see the Customised predictions facility in action see they later chapters of Jones, K and Subramanian, V S (2014) Developing multilevel models for analysing contextuality, heterogeneity and change using MLwiN, Volume 1 , University of Bristol. https://www.researchgate.net/publication/...
by joneskel
Sat Nov 14, 2015 3:45 pm
Forum: MLwiN user forum
Topic: Out-of-sample predictions: Help & guidance
Replies: 2
Views: 4984

Re: Out-of-sample predictions: Help & guidance

I will try and answer but as a lot of this is substantive rather than method /software my answers may be limited. First - customised predictions - these make predictions for the response for chosen specific values of the predictor. This works best when you have built a model using the Add term windo...
by joneskel
Sat Nov 14, 2015 3:05 pm
Forum: MLwiN user forum
Topic: Type III test of fixed effects
Replies: 1
Views: 3198

Re: Type III test of fixed effects

You can test a set of fixed effects simultaneously by a simultaneous Wald test that produces an overall chi-square test. It is illustrated in the Chapter on significant testing in MLwin. See Chapter 6 of https://www.researchgate.net/publication/260771330_Developing_multilevel_models_for_analysing_co...
by joneskel
Tue Jan 06, 2015 4:44 pm
Forum: MLwiN user forum
Topic: GEE approach adjust for clustering in MlwiN
Replies: 1
Views: 3981

Re: GEE approach adjust for clustering in MlwiN

MLwin does not fit a GEE model; it fits a random effects model, that is it estimates an explicit between-group variance You can fit a two level model with people at level 1 nested in cities at level 2 and have a random intercepts model for between city variation. You can then put in a dummy that dis...
by joneskel
Sun Apr 06, 2014 12:28 pm
Forum: MLwiN user forum
Topic: Simple slope test
Replies: 1
Views: 5180

Re: Simple slope test

There are a number of ways that you can evaluate ‘significance’ in MLwin. 1 if the model does not have a discrete response as is estimated by maximum likelihood or restricted maximum likelihood you can do a Likelihood Ratio Test – essentially calculate the difference in the likelihood for the two mo...