Search found 432 matches

by GeorgeLeckie
Mon Apr 01, 2019 1:41 pm
Forum: MLwiN user forum
Topic: Level-2 variance be zero
Replies: 1
Views: 4011

Re: Level-2 variance be zero

Dear Helanidilk, This all sounds fine. Yes, when you set the true cluster variance to be small, in some simulated datasets the estimated cluster variance will be zero. Yes, this is part of the sampling variability of the cluster variance. Note, you should not drop the simulations which give estimate...
by GeorgeLeckie
Tue Dec 18, 2018 4:57 pm
Forum: runmlwin user forum
Topic: Illustrate lower-level interaction by use of a graph similar to margins command
Replies: 2
Views: 5088

Re: Illustrate lower-level interaction by use of a graph similar to margins command

Dear KazimovHH, There is a lot going on here, but at least from a quick scan it looks like you are doing something sensible. You fit your three level model You then set all variables to zero except for the the two variables of interest (one of which is binary, the other continuous) You then predict ...
by GeorgeLeckie
Tue Dec 18, 2018 4:50 pm
Forum: runmlwin user forum
Topic: Cross-level interaction specification in multilevel logit
Replies: 2
Views: 5009

Re: Cross-level interaction specification in multilevel logit

Hi, Yes, your first code and interpretation is correct. In your second code you should include the fixed effect of z3. Including z3 in the random part of the model allows the between-country variance to be a function of country characteristics z3. A heteroskedastic relationship. There is no random s...
by GeorgeLeckie
Thu Nov 29, 2018 11:41 am
Forum: runmlwin user forum
Topic: predict probability and CI
Replies: 3
Views: 5681

Re: predict probability and CI

Hi Rodrigo,

That is correct runmlwin is not comaptabile with margins.

However, you should be able to do what you want simply by using the nlcom command to calculate the expression you have written out.

Best wishes

George
by GeorgeLeckie
Thu Nov 22, 2018 11:01 am
Forum: runmlwin user forum
Topic: p value in mcmc model
Replies: 2
Views: 4888

Re: p value in mcmc model

Dear Rodrigo, The default presented p-value is a Bayesian one-sided p-value (the proportion of the MCMC chain which is the opposite sign to the point estimate). You report a p-value of 0.049 so a conventional two-sided p-value would be more like 0.100 You can actually request the 'properly calculate...
by GeorgeLeckie
Mon Oct 22, 2018 3:44 pm
Forum: runmlwin user forum
Topic: negative binomial example for runmlwin
Replies: 2
Views: 5471

Re: negative binomial example for runmlwin

Dear Ralph, We don't currently have any worked examples, however, here is how you fit the mean dispersion or NB2 version of the negative binomial model in MLwiN for the simple case of one-level data where we have one predictor . runmlwin y cons x, /// level1(id:) /// discrete(distribution(nbinomial)...
by GeorgeLeckie
Tue Oct 09, 2018 4:02 pm
Forum: runmlwin user forum
Topic: Interpretation of the between-country variance as a function of 'x', random slope
Replies: 2
Views: 5447

Re: Interpretation of the between-country variance as a function of 'x', random slope

Negative covariance means that the variance function will be decreasing when x = 0. It will however, increase again eventually at higher values of x and this is what you see.

Similarly, you will see fanning in at x = 0, but will see fanning out again at higher values of x.
by GeorgeLeckie
Thu Sep 27, 2018 4:48 pm
Forum: runmlwin user forum
Topic: Random Slope Model Country Prediction Lines for the Slope
Replies: 2
Views: 5482

Re: Random Slope Model Country Prediction Lines for the Slope

In your plot, look a the y-axis (where x = 0). You should seed that the higher intercepts generally have negative slopes. The lower intercepts generally have positive slopes. This is what we expect given a negative covariance. The stronger the correlation, the more pronounced the pattern. The lines ...
by GeorgeLeckie
Fri May 18, 2018 1:08 pm
Forum: runmlwin user forum
Topic: runmlwin in the growth model
Replies: 1
Views: 3998

Re: runmlwin in the growth model

Dear auctor21,

A conventional three-level linear growth model for a continuous outcome...

. runmlwin outcome cons time, ///
level3(group: cons time) ///
level2(individual: cons time) ///
level1(occasion: cons) ///
nopause

Best wishes

George
by GeorgeLeckie
Tue May 15, 2018 12:27 pm
Forum: runmlwin user forum
Topic: A question about CCMM with runmlwin
Replies: 2
Views: 5117

Re: A question about CCMM with runmlwin

Dear Auctor, The level-1 ID is not actually used by MLwiN, but conceptually the occasion number is the relevant variable to put at the level-1 ID. Note, you will also need to declare the model to be cross-classified, specify the multiple membership strucutre and weights, and fit the model by MCMC. S...