Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feedback from previous #3

Open
8 of 18 tasks
KevCaz opened this issue Feb 29, 2020 · 1 comment
Open
8 of 18 tasks

Feedback from previous #3

KevCaz opened this issue Feb 29, 2020 · 1 comment

Comments

@KevCaz
Copy link
Member

KevCaz commented Feb 29, 2020

  • REALLY Need to add a section about marginal and conditional R2 to this workshop.

  • Plusieurs coquilles/anglicismes, peu d'information sur le calcul d'intervalles de confiance par rapport au ICC, et remplacer les derniers exercices par des calculs de R2 marginal et conditionnel et des tests de ratio de vraissemblance.

  • La présentation .html est très complémentaire et appréciée pour le présentateur, et c'est très bien que toute l'information reste disponible sur le wiki pour les inscrits.

  • All good apart from the bit on ICC and random effects - I did not find this clear and I'm not sure how it links into how the students will use mixed effects models. Remove it, or make clear

  • nous avons ajouté des options supplémentaires au script (sélection de modèles par méthode stepwise et Anova/LRT, fonction vif.mer...) pour donner des alternatives aux participants

  • La création des modèles (M0 à 8 dans le script) donnait un message d'erreur (singular fit). Les modèles fonctionnaient quand même. Je crois que ça doit être dû aux versions de R.studio

  • Je penses qu'il faudrait ajouter certains éléments tirés de l'atelier de Kate Laskowski dans le script, par exemple :
    introduire le concept d'effets aléatoires nichés ou croisés et montrer un exemple de comment l'écrire dans le script.

  • Montrer comment monter un tableau (dans un rapport) rapportant les différents descripteurs du modèle tirés du summary(). (i.e. R2 marginal et conditionnel (avec fonction r.squaredGLMM()), Pentes et ordonnée à l'origine, proportion de variance expliquée par les effets aléatoires, DF, t-value, LLR, p-value (pour les effets fixes)

  • Les participants avaient de la difficulté pour ce qui est de tester les prémisses et l'exploration de données des différents effets. Leur exploration pourrait être simplifée avec les fonctions utilisées dans l'atelier de Kate. (ex. qqplot, qqnorm, dotchart, hist).

  • Slide 12/93 - Typo in the work lake which is missing an l (i.e. it is currently spelt ake)

  • Slide 15/93 - typo: the word ".alert [" slipped into the first sentence on this slide and there is an extra ] at the end of that sentence; last bullet point is missing the word "we" after "but"

  • Slide 17/93 - You cannot say "we will meet those terms frequently" in English in this context, should instead be something like "In the LMM literature you frequently encounter the terms 'fixed' and 'random' effects". The slide later says "Challengenitions", this should be replaced by "Definitions"

  • Slide 20/93 - says "effet aléatoire" which should be random effect

  • Slide 21/93 - Incomplete first sentence. add a period after and skip a line after the sentence about the intercepts being from a normal distribution.

  • Slide 22/93 - Same comment as above - first sentence is not complete

  • Slide 23/93 - add "it is" before the first "just" to make this sentence complete. Add a period after "visualize" and skip a line.

  • Slide 26/93 - Adjust text under the low ICC figure "because little correlated" does not make sense.

  • Slide 31/93 - The equation is a missing an equal sign (or a function of sign)

@quantitative-ecologist
Copy link
Member

quantitative-ecologist commented Mar 3, 2021

Hi all,

I modified the ICC part a bit so that it represents more what ecologists do e.g. when they want to know the proportion of variance explained by a random effect. I included the formula so people can see how to calculate the ICC. I also changed the graphs so that people see an example with differences in intercept variance which make more sense with the content presented in the intro.

I think a part on conditional and marginal R2 would fit well after this section as @KevCaz suggested.

katherinehebert added a commit that referenced this issue Mar 23, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants