Comments on: R-Sessions 16: Multilevel Model Specification (lme4)

By: Rense Nieuwenhuis

Rense Nieuwenhuis — Sun, 15 Nov 2015 07:41:21 +0000

Thanks!

A .PDF version of most of the blogs about R is available: http://www.rensenieuwenhuis.nl/documents/Applied%20R.pdf

By: John Fresen

John Fresen — Sat, 14 Nov 2015 08:58:33 +0000

Excellent discussion of multi level model fitting in R.
Would you be kind enough to supply a PDF copy so that I could give a hand out to students.
Many thanks

John Fresen

By: Rense Nieuwenhuis

Rense Nieuwenhuis — Thu, 03 Sep 2015 14:20:19 +0000

Absolutely!

By: SALLY

SALLY — Mon, 27 Jul 2015 14:14:28 +0000

That’s a great post! I have one question…could we include a Level 2 variable as DV e.g. school review scores in your example?
lmer(schoolreviewscore ~ standLRT * schavg + (1 + standLRT | school), data=Exam)

By: marcusmorrisey

marcusmorrisey — Tue, 16 Jun 2015 19:11:10 +0000

Great post, very clear thanks!

By: Rense Nieuwenhuis

Rense Nieuwenhuis — Tue, 21 Apr 2015 22:14:29 +0000

Hi Laura,

thank you for your comment, and for your interest on my (old) blogpost. I refrained from interpreting the results, as interpretations are really dependent on the purpose of the regression model. All parameter estimates are important, but which ones you emphasise in your interpretation depends on your research questions / hypotheses / modelling goals.

Should you have more specific questions, you are always welcome to contact me.

By: Laura

Laura — Mon, 20 Apr 2015 17:55:38 +0000

I found this helpful to illustrate the different syntaxes but could you please provide some information on the interpretation of the output? And also how it differs across scenarios? What exactly are we meant to be looking at to determine our coefficients of interest?

By: Joe Powers

Joe Powers — Wed, 30 Jul 2014 21:12:54 +0000

Second that. This is the clearest answer I’ve found in 2 years of trying to wrap my head around this. Thanks.

By: d l rogers

d l rogers — Sat, 22 Mar 2014 21:15:31 +0000

Two years later… me too! I need to work with several 2- or 3-level models, and can’t seem to find any instructions for how to specify those in lme4 or nlme.

By: robie

robie — Wed, 08 Aug 2012 11:42:45 +0000

great – very clear and good to include the crossed design. be good to see SPSS syntax equivalents and often people start with spss and move onto R