Search found 49 matches

by Harvey Goldstein
Tue Jan 29, 2013 9:03 pm
Forum: Realcom user forum
Topic: Problem specifying weights
Replies: 5
Views: 7171

Re: Problem specifying weights

Yes -weights should certainly be used in model of interest. MLwiN will allow this.
It is often not serious if you do not use weights in the imputation.
Harvey Goldstein
by Harvey Goldstein
Mon Jan 21, 2013 3:21 pm
Forum: Realcom user forum
Topic: Imputation model
Replies: 2
Views: 4275

Re: Imputation model

The equations displayed are for the complete imputation model that you have specified. See notes on realcom for an example.
Harvey Goldstein
by Harvey Goldstein
Mon Jan 21, 2013 3:17 pm
Forum: MLwiN user forum
Topic: missing data
Replies: 1
Views: 3338

Re: missing data

Mlwin will by default always use casewise deletion for missing data - even when carrying out a Bayesian fit using MCMC. REALCOM was written to overcome this and it links with mlwin - but will handle only 2 levels at present. You could impute using 24 dummies for country. Go to realcom web site to se...
by Harvey Goldstein
Mon Jan 21, 2013 3:12 pm
Forum: Realcom user forum
Topic: Problem specifying weights
Replies: 5
Views: 7171

Re: Problem specifying weights

The weights facility is still not implemented. Recent work by Carpenter has addressed this issue - see new book by Carpenter and Kenward (Wiley, 2013)
by Harvey Goldstein
Mon Jan 21, 2013 3:10 pm
Forum: Realcom user forum
Topic: Become insignificant after MI
Replies: 1
Views: 3245

Re: Become insignificant after MI

This is often what you might get when you do imputations - suggests that the complete cases analysis produced biased results.
Harvey Goldstein
by Harvey Goldstein
Mon Nov 05, 2012 11:41 pm
Forum: MLwiN user forum
Topic: Level 2 residuals after using multiple imputation
Replies: 2
Views: 4039

Re: Level 2 residuals after using multiple imputation

Mlwin will not automatically combine residuals. You will need to do this manually. We will look into doing this.
Harvey Goldstein
by Harvey Goldstein
Wed Oct 17, 2012 10:01 am
Forum: Realcom user forum
Topic: Imputation Model
Replies: 2
Views: 3849

Re: Imputation Model

You should generally use all variables that will appear in the model of ineterst. In addition you can condition on any auxiliary variables that might be useful for satisfying the missing at random assumption.
Harvey Goldstein
by Harvey Goldstein
Mon Sep 24, 2012 2:45 pm
Forum: Realcom user forum
Topic: Interaction Effects Model
Replies: 2
Views: 3668

Re: Interaction Effects Model

An interesting question. At the moment the MOI should not contain interaction terms. There are different approaches in the literature to incorporating interactions (including polynomials) - see e.g. van Buuren S (2007). Multiple imputation of discrete and continuous data by fully conditional specifi...
by Harvey Goldstein
Mon Sep 24, 2012 2:37 pm
Forum: Realcom user forum
Topic: Realcomimputeload error: `var’ found where numeric variable
Replies: 5
Views: 6176

Re: Realcomimputeload error: `var’ found where numeric varia

I'm puzzled by this. I don't think that realcom will ever write 'NA' in these files. It is just possible that if the model fails to converge you could get 'NAN'.
Harvey Goldstein
by Harvey Goldstein
Mon Sep 24, 2012 2:29 pm
Forum: Realcom user forum
Topic: Multivariate Ordered regression: error matrix dimensions
Replies: 6
Views: 6697

Re: Multivariate Ordered regression: error matrix dimensions

Re your second question categorical expl variables should be entered with dummy variable coding.
Re first Q try running as a single level model. Can you also confirm which are responses and which explanatory.
Harvey Goldstein