filter_out is so simple of an upgrade but so nice for readability
filter_out is so simple of an upgrade but so nice for readability
Well, I'd challenge the claim that biology has consistently better practices in that regard. The requirements for raw data and script are also quite loose, journal and sub-field dependent (although it's getting better!). You're pointing to an inter-disciplinary replication crisis, really!
A super useful and comprehensive summary!
#ReproducibleCode
Of course you did! Thanks for sharing.
Thanks Andrew. I love your blog!
I'll risk tagging folks I've seen mention the book @aammd.bsky.social @andrew.heiss.phd @elucidating.extradimensional.space @noamross.net @marikendewit.bsky.social
Which makes me think, does anyone have tips on how to run a successful book club on a textbook like Statistical Rethinking?
We'll be meeting every two weeks, following Richard's older schedule at half speed. Different folks would sign up to lead the discussion. Mix of faculty and grad students.
Next semester I'm organizing a Statistical Rethinking book club for the bio dpt at @carleton.ca
Interestingly we also decided to progress at half speed. This new 2 sections idea will be perfect if I ever decide to run future iterations of that book club (gotta learn the material first though!)
"What’s unfolding now is more than dishonesty—it’s the unraveling of any shared understanding of what education is for"
"The CSU isn’t investing in education—it’s outsourcing it, paying premium prices for a chatbot many students were already using for free."
Wikipedia article for barber-surgeon. The barber surgeon was one of the most common European medical practitioners of the Middle Ages, generally charged with caring for soldiers during and after battle. In this era, surgery was seldom conducted by physicians. Instead, barbers, who possessed razors and dexterity, were responsible for tasks ranging from cutting hair to pulling teeth to amputating limbs.
The scientist-statistician was one of the most common professions in 21st century universities. In this era, statistics was seldom conducted by trained statisticians. Instead, subject experts, who possessed hard- and software, were responsible for tasks ranging from study design to causal inference
Split image: Main photo — black-and-white portrait of young Jocelyn Bell Burnell in the 1960s, smiling broadly in front of the large lattice structure at Mullard Radio Astronomy Observatory (SW of Cambridge in the UK). Inset (left) — color portrait of Dame Jocelyn Bell Burnell today, short graying hair, glasses, and a warm smile, wearing a blue jacket.
#OTD in 1967, Jocelyn Bell discovered pulsars & changed our view of the universe. 𝘉𝘶𝘵...
The #Nobel Prize in #Physics for their discovery went to to her thesis supervisor Antony Hewish & to astronomer Martin Ryle.
theconversation.com/fifty-years-...
#WomenInSTEM #astrophysics #MatildaEffect
I love the JBB story for a few reasons:
- stringing 120 miles of wire
- months of analzying paper roll data
- her PI was like "that can't be right, do it again," the motto of PI's everywhere.
- rare case of a landmark discovery becoming ubiquitous in popular culture for no particular reason
Let me guess. Only the French... Erik Satie?
We have bison farms in Quebec
There is some evidence that dentists struggle with mental health quite a bit
www.nature.com/articles/s41...
Look up @carlbergstrom.com course: "modern oracles or bullshit machines". It should help with that line of thinking.
Visionary
The lesson resonated with me so strongly when I listened to Richard's lecture last year, I've been waiting for a relevant occasion to share the gospel
m.youtube.com/watch?v=8qzV...
A lecture slide titled "science as amateur software development" and subtitled "a lesson shamelessly pilfered from McElreath 2023". The background of the slide is a painting by David Teniers entitled "An alchemist in His laboratory".
A lecture slides with two memes illustrating the point written at the bottom of the slide: "a lot of contemporary, western science requires software development". The meme on the left shows a person baiting a squirrel but hiding a knife. The squirrel represents ecology students, the bait is biology and the knife is programming and statistics. The meme on the left shows captain Jack Sparrow (the main protagonist in the pirates of the Caribbean movie franchise) licking a rock and making a disgusted face. Jack Sparrow is a stand-in for the students and the rock represents RStudio.
Tomorrow I'll have the pleasure of teaching version control to @carleton.ca ecology undergrads
As a preamble, to drive the importance of professional tools for doing professional science, I've stolen @rmcelreath.bsky.social's entire lesson on the topic
Of course, there will be classical memes
Marvel: "infinity war is the most ambitious crossover event in history"
McElreath, Gelman & Vehtari:
I had to check what I recommend in my book. And Poison priors (right) are also cursed. I'm working on something new with Gelman and Vehtari, so will make a note to review our examples for some consistency in these contexts.
Yup, I've found this same need for periodic re-teaching of those fundamental concepts (both to myself and to others).
I spent an hour yesterday reinforcing to a student the idea that smaller p-values != stronger effects, and that p-value = P(data | NH true) and not P(NH true | data).
Very nice article. Will definitely reuse that example when teaching undergrads.
Very much looking forward to this, and more than happy to support a good cause
That seems so unbelievable to me that this is happening. What's with the obsession that talent is best found *elsewhere*?