Talk:Generalized linear mixed model

How to select a GLMM
The most commonly-used method of model selection is via AIC. It has been tricky, though, to use AIC with GLMMs. The recent paper by Saefken et al. gives a practical method for working with AIC. Thus, Saefken et al. advance our abilities to select the best GLMM. For more details, see the paper (it is free online, via the doi). 31.55.87.180 (talk) 17:27, 24 February 2015 (UTC)


 * Please add text to the article that makes this reference relevant. Your fervor in adding this references against accusations of refspam makes me suspicious.  I don't know if you are the author of that paper, but please take note of WP:COI anyway. Tayste (edits) 22:12, 24 February 2015 (UTC)
 * PS: beware of edit wars that will get you blocked. Tayste (edits) 22:14, 24 February 2015 (UTC)
 * You are right that I am a little fervent in wanting to add the reference. I am not an author though; nor do I know any of the authors.  Rather, I like to support using AIC.  The reference seems to extend the cases in which AIC is easily used.
 * I have added some text to the article, as you request. I am not an expert on GLMMs though; so I am hoping that the new text is okay.   SolidPhase (talk) 20:43, 26 February 2015 (UTC)

Source for improvements to this article
I think this article is worth expanding. This DRAFT r-sig-mixed-models FAQ has a lot of information and many useful references. Tayste (edits) 21:09, 26 February 2015 (UTC)
 * Agreed. Perhaps this should be in a (new) External Links section? SolidPhase (talk) 09:36, 3 March 2015 (UTC)

Format for Notes and References
This article previously had separate sections for Notes and References. That was changed: now there is just References. The change makes it impossible to cite different pages (in different notes) within the same book. For an example of how to do such citation, see e.g. the article Akaike information criterion. The individual notes can be set up so that they link to the References (as in the AIC article). I believe that it would be better to have this article use the same style of Notes & References as the AIC article. SolidPhase (talk) 09:47, 3 March 2015 (UTC)

Meaning of symbols
I have noticed that the article provides no explanation to what the Phi and Theta symbols mean in the equations. As someone unfamiliar with these models, I could only find Theta in the article about generalized linear models for parameterization, but not Phi. --2A02:8109:9EC0:180A:F02E:A404:EDFA:37B0 (talk) 20:13, 26 July 2020 (UTC)


 * I simplified the definition a bit by removing the explicit formula for the exponential family, thereby removing the phi and theta. Instead I now just let the reader know that this short definition of the generalised linear mixed model can only be understood after looking at generalized linear model and mixed model. Gustav (talk) 17:58, 25 April 2021 (UTC)