Logistic Regression in R - With Flexplot
ฝัง
- เผยแพร่เมื่อ 29 ก.ย. 2024
- See my original video on GLMS here: • Become a generalized l...
Sensitivity/Specificity/PPV/NPV Explanation: geekymedics.co...
Paper about ambiguity (btw, this is a first draft, so it's VERY rough): quantpsych.net...
Learning Objectives:
#1. Understand how to perform logistic regression in R.
#2. Know how to visualize logistic regression models in flexplot.
#3. Know how to visually and statistically compare two logistic regression models.
This video is part of my multivariate playlist: • Multivariate Statistics
And here's a paper I wrote about my eight step approach to data analysis: psyarxiv.com/r...
Undergraduate curriculum playlist (GLM-based approach): www.youtube.co....
Graduate curriculum playlist (also GLM-based approach): www.youtube.co....
Exonerating EDA paper: psyarxiv.com/5...
Download JASP (and visual modeling module): www.jasp-stat.org
Man, incredible content! Thank you very much. I'm an mechanical engineer trying to do statistics and this is helping SO much. You deserve and will have more views in the future, keep it up.
Would be great to have access to the other GLIM regression examples. BTW very nice synthetic functions in Flexplot
It's actually marvelous0 that you used kinda "real" data and showcased that sometimes a certain model may fit the data extremely poorly.
Careful, when you start you can't stop binging his content!
Caution to be especially advised after coffee
Why not a poison or negative binomial distribution? Since you’ve got # minutes (aka time, continuous) and a discrete observation (dead, alive) then couldn’t a poisson or neg binomial distribution also make sense?
I think it is because the variable being modelled has a bernoulli distribution (dead, alive). In addition, both poisson and negative binomial only have integer values, so I don't think they are appropriate for modelling continuously measured time.
how to calculate statistical significance if we have a hypothesis? Like willpower is associated with died/not died
After a few people died in the battle, the snap made 50% of the survivors disappear.
Ha!
How would we incorporate a random variable in the full and reduced model, as well as the model comparison (line 26)? Something like this?
full_model