Mixed Model Notation - A Simple Explanation

[@alifarahani4398]

This guy is AMAZING. The best video you can find on the internet on this concept.

[@erdeduft]

After my frustration yesterday in the statistic course for my PhD, i am so relieved and happy to have watched this video!! Pure CLARITY that i gain about this subject. This Model ist important for my study thesis, and i really needed this! Many thanks!😇

[@QuantPsych]

Glad it helped! I do have more updated videos that I recently published. You can also take a live or on-demand class from me at simplistics.net!

[@itamar.j.rachailovich]

15:10 , I think you meant that "-1" is the way we tell R that the intercept should not be random, but only the slopes, however you said: "-1 is the way we tell R that the slopes should not be random.

[@QuantPsych]

correct

[@khosbeh5467]

GREAT work. Fantastic presentation. You make complicated concepts so understandable. I have been an instructor for 37 years and I admire your abilities and execution. KEEP UP the great work.

[@QuantPsych]

Thanks!

[@seaz9992]

These videos are so well done and informative. I took MLM a bit ago and forgot some of the notation, this was very helpful!

[@lucagiraldi372]

Amazing video! at 15:10, when you talk about the "-1", this is to indicate that the intercept should not be random, not that the slope should not be random, isn't it?

[@rHk217]

I was thinking the same thing.

[@tahamansoor6548]

Awesome!! This was so complicated for me before this video. You made it so easy. Thanks!

[@CanDoSo_org]

Why I got a different plot at 10:55. There is only one red line, no green line.

[@161miracle]

me too. did you figure out why?

[@jd.1184]

@@161miracle me too

[@Robin-wo8gj]

Thank you so much Dustin ! Your videos are a lifesaver.

[@fazlfazl2346]

@0.36 I have a question. We usually have two parts for regression. One is before fitting the model and the second is prediction using the fitted model. What is confusing for me is that (Depression_i = B0_i + B1*Stress_i) seems to be something that we will use during prediction because at prediction we are interested in a single individual. While fitting the model we take all individuals together to fit the regression line so we do not think in terms of single individuals (the ith individual) but all the individuals. So @0.36 seconds, what part of the model are you referring to? If we are fitting the regression line using OLS, how does the subscript I matter while fitting using OLS? Please clarify.

[@thelateknights]

Is there a certain number of levels that should be present to use a mixed model? Like if we look at gender assigned at birth, there would be 2 levels, would it be appropriate to include that in a mixed model, or just dummy code that?

[@giovannifontanetto9604]

I hope to be able to explain things like you one day

[@itamar.j.rachailovich]

Thanks for your great videos. You are funny as well. Some how the plot i got from flexplot function is different than yours: there was no line for the random effects (and no legend as well).

[@vincentcabibel9838]

Hi, very great as usual ! I'm trying to use flexplot to visualize my lmer models however I consistently get a "$ operator not defined for this S4 class" error, even when using your code (without any change) at 10:30. Any idea of what's happening ?

[@QuantPsych]

When was the last time you updated flexplot? I think in the Fall, I fixed all those bugs. Try installing from github.

[@vincentcabibel9838]

@@QuantPsych Installed right before using few days ago, but I tried on another computer and it worked. Seems there is an issue with the computer then. Thanks for the great package !

[@kashyappandya6357]

As per explanation at 5.15 of video, for intercept : that people who come to respective therapist have already average score of depression is 11 and for slope : where average or general slope is +0.5, but in given group slope is -0.25, which shows efficiency or effectiveness of therapy(for example, may be), which is low compare to average slope across group. am I right? please somebody answer...

[@nghieply7951]

Thank you for the video. This is what i need to understand about LMEM

[@lotto.2618]

I’ve got this model from my professor: Yij=B0j + B1jxij + eij Boj= y00 + u0j B1j= y10 + u1j What does the B1jxij or inserted Y10xij +u1j mean?

[@romainouldammar3660]

Very good video! Thank you. In the followings models (Alcuse~1+AGE_14+(1|ID ); Alcuse~1+AGE_14+(AGE_14|ID)), does the random intercept (for both models) correspond to the variation of alcohol consumption at Age 14 ? If yes, could you explain why the estimation differs (0.5966 vs 0.6355)?

[@LeoWei-dd9to]

Your video is so helpful! Thank you!

[@lenag3329]

awesome explanation, thank u!

[@brandyhorne1997]

You are awesome - thank you!

[@pairotechakranon1568]

Dear Prof. For link vdo part 1 is missed, can u fix it? this is the best stat channel that I love :)

[@QuantPsych]

Oops! This link should work: kzbin.info/www/bejne/mZDXio2uhrGieqM

[@jorgemmmmteixeira]

Is the code and dataset available somewherE? tks

[@QuantPsych]

You can run my code as is if you install flexplot.

[@fjumi3652]

Thank you for this video! :D :D :D

[@denisfitzpatrick6781]

The background music is distracting.