Happy to share that my first solo-authored paper is now published in the Annual Review of Statistics and Its Application.
📑 Reproducibility in the Classroom
🔗 www.annualreviews.org/content/jour...
Speakers include
Matt Castle, @calexander91.bsky.social, @forrt.bsky.social, @minedogucu.com, @minecr.bsky.social
2/2
On 12th March 2025 (3pm-5pm, UK time), @rss-teach.bsky.social and TALMO will be hosting an online event on *Teaching Open, Transparent, and Reproducible Data Analysis*. See talmo.uk/2025/dataana....
Free and open to all.
1/2
As Thom says, power is strictly a frequentist concept, but there still is a relatively big literature on sample size determination in Bayesian context and so Bayesian power analysis is "a thing" in practice. Kruschke's 2013 is a good place to start: jkkweb.sitehost.iu.edu/articles/Kru...
New updated slide deck of my talk "rage against the machine learning"
www.slideshare.net/slideshow/ra...
{effectsize} version 1.0.0 has been released! 🍕👹
Not many changes in this update, but his is quite a milestone 🥳 - the second stable package from @easystats.bsky.social, with more to come soon!
#rstats
easystats.github.io/effectsize/
We are excited to announce pal: a package of ergonomic #LLM assistants designed to help you complete repetitive tasks quickly in RStudio or Positron!
Created by @simonpcouch.com, the documentation walks through built-in pals and how to make your own: simonpcouch.github.io/pal/
#RStats
The BPS Statistics and Research Methods Advisory Panel is a panel of academic researchers who advise editors and provide stats or research methods related peer review for BPS journals.
The Panel is currently recruiting for new members.
www.bps.org.uk/bps-statisti...
For those who love LLMs, especially open-source LLMs, here is a very useful and up to date list:
github.com/allenai/awes...
The source I now most frequently refer to in my peer reviews: datamgmtinedresearch.com/document#doc...
Documenting things is hard but people really can't check/use your work if it doesn't have some docs. We can do better!
Interact with Bluskey in R using the atrrr package. E.g, get your followers and followees:
library(atrrr)
my_handle <- 'username'
auth(user = my_handle, password = 'password')
my_followers <- get_followers(actor = my_handle, limit = Inf)
my_followees <- get_follows(actor = my_handle, limit = Inf)
S7 0.2.0 is now on CRAN, thanks primarily to the hard work of @t-kalinowski.bsky.social. S7 is a collaborative effort to supersede S3 and S4 and you can read all about it (and what's new in 0.2.0) at www.tidyverse.org/blog/2024/11... #rstats
You can open any high quality medical journal and read authors talking about "trends" in estimates, almost always when the attached p-value is just outside the "significance" cutoff. This kind of thing is maddening, because it simply doesn't reflect reality. It's not just wrong, it's...stupid 😬. 1/
