🔥 🔥 🔥 😱
https://www.smbc-comics.com/comic/p-3

#Pvalues #Research #Comics

"Abandon statistical significance, all ye who enter here"

I really need a sign for my office with this slogan.

Until then, I'll leave this #OpenAccess paper by McShane et al here for you to peruse: https://www.tandfonline.com/doi/full/10.1080/00031305.2018.1527253

#Stats #Statistics #StatisticalSignificance #PValues #Rstats

Large-scale examination of differential RNA expression study datasets submitted to the @NCBI #GEO repository from 2008 to 2020 reveals widespread #bias, as assessed by distribution of #pvalues & estimated proportions of true null hypotheses #PLOSBiology https://plos.io/3IJHMye

Large-scale examination of differential RNA expression study datasets submitted to the @NCBI #GEO repository from 2008 to 2020 reveals widespread #bias, as assessed by distribution of #pvalues & estimated proportions of true null hypotheses #PLOSBiology https://plos.io/3IJHMye

Large-scale examination of differential RNA expression study datasets submitted to the @NCBI #GEO repository from 2008 to 2020 reveals widespread #bias, as assessed by distribution of #pvalues & estimated proportions of true null hypotheses #PLOSBiology https://plos.io/3IJHMye

#Penises & #Pvalues

Erect length increased significantly over time (QM=4.49, df=2, p=0.04)

What I want to know, however, is how the size of penises is related to the size of p-values. Do p-hacker have smaller peckers?

https://pubmed.ncbi.nlm.nih.gov/36792094/

There's a #debate in #ExperimentalPhilosophy (#xPhi) about whether #intuitions about philosophical thought experiments are "stable" (vs. manipulable).

Some (e.g., @xphilosophy) seem convinced of stability—e.g., because some replications of instability results and new experiments find #pValues > 0.05.

Alexander & @jonweinberg express a different view (https://philpapers.org/rec/ALEPMP). I agree that the debate needs parties to agree on a smallest effect size of interest.

Page images attached.

Talk to the UCL R Users group about p values and our web app. (an R shiny app):

Now up on YouTube
youtube.com/watch?v=tFTc_k…
#statistics #shiny #pvalues #falsepositives #UCL

@lakens @raoulvanoosten
Talk to the UCL R Users group about p values and our web app. (an R shiny app).
Now up on YouTube
https://www.youtube.com/watch?v=tFTc_kQ9kSM
#statistics #shiny #pvalues #falsepositives

Talk to the UCL R Users group about p values and our web app. (an R shiny app).
Now up on YouTube
https://www.youtube.com/watch?v=tFTc_kQ9kSM
#statistics #shiny #pvalues #falsepositives

Talk to the UCL R Users group about p values and our web app. (an R shiny app).
Now up on YouTube
https://www.youtube.com/watch?v=tFTc_kQ9kSM
#statistics #shiny #pvalues #falsepositives

Talk to the UCL R Users group about p values and our web app. (an R shiny app): http://fpr-calc.ucl.ac.uk/
Now up on YouTube
https://www.youtube.com/watch?v=tFTc_kQ9kSM
#statistics #shiny #pvalues #falsepositives

Does anyone know a good paper/article/video that explains that (and why) the p value follows a uniform distribution when the null is true (i.e., when there is no association/effect in the population)? @lakens (please correct me if I got this wrong) #pvalue #pvalues #stats #statistics