Page 1 of 1
Multiple Memebership & 3-Level model analysis
Posted: Wed Nov 16, 2016 7:20 am
by auctor21
I'm new to MLwiN and want to use MLwiN to analyse my longitudinal data (multilevel- 3 levels). The data structure takes the form of repeated measures (level 1) nested within the individual (level 2) and individuals nested within higher level cluster (level 3). But, I have multiple membership in level 3.
Namely, first level is time, second level is students and third level is school.
Is there any way to an analysis of these structures?
Can you guide the method for analysis?
Re: Multiple Memebership & 3-Level model analysis
Posted: Wed Nov 16, 2016 4:01 pm
by ChrisCharlton
Multiple membership models are covered in chapter 16 of the MCMC guide (see
http://www.bristol.ac.uk/cmm/media/soft ... mc-web.pdf), so I would suggest working through this initially if you haven't already done so.
Re: Multiple Memebership & 3-Level model analysis
Posted: Fri Nov 25, 2016 6:43 am
by auctor21
Through your work, I was able to confirm the basic methods.
Also, by checking out LEMMA, I had the opportunity to confront various methods.
However, I was unable to obtain the core information that I was looking for.
Like mentioned earlier, I am trying to to a 3-level analysis with 1-level being time, 2-level being students, and 3-level being the school.
Here, the school becomes the multiple membership.
Also, level 2 and 3 are included in the explanatory variable as well.
I was wondering if you would be willing to provide a method to analyze this model in more detail.
Through my research, I was unable to figure out the solution.
(for example, in order to solve 3-level, multiple membership, and repeated measures I looked at
Module 15: Multilevel Models, Module 13: Multiple Membership Multilevel Models
, Module 15: Multilevel Modelling of Repeated Measures Data
But I was unable to find the solution regarding the model I am dealing with.
Your advice would be greatly appreciated.
Thank you.
Re: Multiple Memebership & 3-Level model analysis
Posted: Mon Nov 28, 2016 2:48 pm
by ChrisCharlton
Although they don't describe your specific model, the basic elements of the resources that you refer to can be used together to build the more complex model. It is also sometimes useful to look for a published paper that fits similar models so that you can use this as a guide. If you have problems understanding specific parts of the module I would suggest contacting the relevant authors of the materials.