Search found 112 matches

by billb
Wed Jan 24, 2018 1:40 pm
Forum: R2MLwiN user forum
Topic: Relatively simple MCMC logistic model needs many iterations
Replies: 3
Views: 1504

Re: Relatively simple MCMC logistic model needs many iterations

Hi Gerine, This is an interesting email and sounds like you may have specific features of your dataset that are causing these convergence difficulties. I'd happily take a look if it helps? A 3-level model isn't what I'd define as 'relatively simple' and it may be that you do not have much if any var...
by billb
Thu Nov 30, 2017 1:01 pm
Forum: MLwiN user forum
Topic: Power/Effective Sample Size
Replies: 2
Views: 2142

Re: Power/Effective Sample Size

Hi Jillian, I think you have got yourself completely confused here due to the use of the term 'effective sample size' for two totally unrelated concepts. In MCMC estimation, being a simulation-based procedure, the user has to run the technique for a large number of iterations to get estimates that a...
by billb
Tue Nov 28, 2017 12:27 pm
Forum: MLwiN user forum
Topic: Level 2 variance increases with inclusion of Level 1 predictors
Replies: 2
Views: 1265

Re: Level 2 variance increases with inclusion of Level 1 predictors

Hi Paris, This is a quite common phenomenon and easy to explain. Imagine for example growth curves between individuals measured yearly for 10 years. In a model without age we would have lots of variation within an individual as their heights at age 4 say will be very different from their heights at ...
by billb
Fri Nov 24, 2017 7:44 pm
Forum: MLwiN user forum
Topic: question about small number group unit. NEED HELP! THANKS!
Replies: 1
Views: 721

Re: question about small number group unit. NEED HELP! THANKS!

Hi, You should always build up your model step by step starting with random intercepts and pressing more when you go to random slopes instead of start to help potential convergence. Convergence is more of an issue if you smaller numbers of level 2 units and particularly if there is little variation ...
by billb
Wed Sep 06, 2017 8:44 am
Forum: MLwiN user forum
Topic: Binary MLM assumptions
Replies: 1
Views: 1205

Re: Binary MLM assumptions

Hi Jessica, In answer to your questions: a) Yes at higher levels you can check the normality assumption via residual plots - the residuals at level 1 should not be normal b) Not sure what you want to do here - you can plot the transformed fitted line to your data for example or you could test a quad...
by billb
Tue Sep 05, 2017 3:16 pm
Forum: MLwiN user forum
Topic: Informative priors
Replies: 5
Views: 1952

Re: Informative priors

No worries Ron,
Glad to be of help - basically the larger the clustering variance the harder it is to justify fitting a multilevel model rather than simply collapsing the data by aggregating up a level.
Best wishes,
Bill.
by billb
Mon Sep 04, 2017 12:19 pm
Forum: MLwiN user forum
Topic: Informative priors
Replies: 5
Views: 1952

Re: Informative priors

Hi Rdmcdowell, Thanks for the clarification. Good to see people using my MCMC book :) The first thing to note is that a variance of 4.5 is crazily big in a logistic regression model (particularly if you are getting this from the default 1st MQL) and would usually indicate that most clusters are all ...
by billb
Mon Sep 04, 2017 9:31 am
Forum: MLwiN user forum
Topic: Informative priors
Replies: 5
Views: 1952

Re: Informative priors

Hi Rdmcdowell, When you say do not work well do you mean simply that the chains mix poorly? My only experience here really is in my paper Browne et al. (2007) in Statistical Modelling - https://scholar.google.com/citations?view_op=view_citation&hl=en&user=mxNZAmoAAAAJ&cstart=20&pagesize=80&citation_...
by billb
Wed Aug 30, 2017 7:49 am
Forum: R2MLwiN user forum
Topic: Interpretation of coefficient in multilevel cumulative logit model
Replies: 4
Views: 2162

Re: Interpretation of coefficient in multilevel cumulative logit model

No worries both. Basically if the probability of being in categories k and upwards is p then the probability of being in categories k-1 and downwards will be 1-p and if logit(p) = a then logit(1-p) = -a hence when you switch the base category then the main change will be a switching of signs for all...
by billb
Wed Aug 30, 2017 7:44 am
Forum: runmlwin user forum
Topic: Advice on cross-classified models with large data sets
Replies: 11
Views: 3377

Re: Advice on cross-classified models with large data sets

Hi rdmcdowell,
Am afraid I didn't write runmlwin so can't help here though I'll mention this to George Leckie and Chris Charlton to reply to you,
Best wishes,
Bill.