Explaining Variance Explained in Mixed Models

[@FilippoGianburrasca]

8:40 why it doesn't report also the improvement in the slope variance (i.e. how each cluster slope differs from the grand slope)?

[@FilippoGianburrasca]

7:53 you ran the model at line 56 but then you commented it like it was the model at line 60, which you didn't run (if I'm not mistaken)

[@elliothoward-spink2120]

model.comparison() from the RRPP package wont let me use an lmer object - do you know why? it says 'Objects must be lm.rrpp fits' :-(

[@taranaferdous2858]

I do not see the felxplot in r package. Is it removed?

[@QuantPsych]

No, you have to install from github.

[@dann4179]

What does it mean the ghost.line black line in the compare.fits you are doing ?

[@rachitsingh98]

Thanks for the nice explanation, Dustin! When I tried to follow this approach I was not shown the R-squared change values. Could this be because I was using the "glmer.nb" function and that the R2 calculation is tricky in that case?

[@pranavjain9799]

Thank you so much. Teachers often just skip over what "variance explained" mean. Also, at 1:58, I think you mean't to say "Because they are heavier"

[@FilippoGianburrasca]

* fatter

[@firuzerehimova6355]

you are a supeeer annoying but also addictive teacher thanks x

[@QuantPsych]

umm....thanks?

[@nazirullah8365]

do you any knowledge regarding i need help

[@vietbio1]

Hi, I got Error "$ operator not defined for this S4 class" when I try compare.fits function. How can I fix that?

[@julitwiga]

I'm running RStudio 2021.09.2 Build 382 "Ghost Orchid" Release for Mac OS, and flexplot 0.13.3 and I get the following message when I run the baseline you mention in this video: Error in reformulate(paste0("(", vapply(findbars(f), deparse1, ""), ")"), : unused argument (env = environment(f)). Any thoughts?

[@eyadha1]

Thank you very much. Can you make a video abot sample size in mixed effects models? thank you

[@helenadamson6709]

Your videos are so helpful - and enjoyable - but ... " package ‘flexplot’ is not available (for R version 4.0.2)"

[@QuantPsych]

It must be installed from github: install.packages("devtools") devtools::install_github("dustinfife/flexplot")

[@zimmejoc]

Hi all. I'm running Flexplot version 0.19.1 so quite a bit newer than what Dustin is running here. If you DO NOT have the r squared change output, you'll need to detach the lmerTest package. At least when I detached lmerTest, I got the r squared change output.

[@tatjanajak]

I can't load data(math) in R. It gives the message: object 'math' not found.

[@tatjanajak]

Sorry, i take this back. The math data is the part of the flexplot project. Just needed to install it properly.

[@QuantPsych]

Glad you figured it out!

[@Long_Live_Freedom_Damn_It]

Voice is not syncronized

[@ajaydhungana1921]

Hi..can you make short video on how REML estimates variance components in mixed model.. i mean whats goes inside of REML. hoe does it treats the data sets..Thanks...SUbscribed your channel.

[@rachelkeszycki6838]

Hi, when I use model.comparison for this example, it prints everything except for the $r_squared_change section. Do you know why this may be?

[@QuantPsych]

Sorry! It's fixed in the most recent iteration of flexplot. devtools::install_github("dustinfife/flexplot", ref="development")

[@yutingcui2485]

@@QuantPsych So sorry, I got the same print in using model.comparison function, the results didn't show R2 change. How do solve it?

[@dann4179]

What does it mean the ghost.line black line in the compare.fits you are doing ?