James Steele

@jamessteeleii.bsky.social

I'm just a guy who's an academic for fun

I'd previously shared a preprint summary of a talk I gave last year on meta-analytic approaches to exploring intervention effect heterogeneity in resistance training. But, based on feedback (and an invitation to submit something for a special issue), I've now shaped it up into a proper paper... 1/8

If you aren't sat on a garden bench enjoying a British summer evening, with Kiki on your lap, wearing yellow crocs with socks and a Comedian t-shirt, and eating an ice lolly... what are you even doing with your life?

A person relaxes on a wooden garden bench with a tabby cat curled on their lap while eating a red ice lolly. The person gazes ahead as the cat looks directly toward the camera. The bench sits beside a garden bed of leafy plants, with a lawn in the foreground and a wooden fence and shed lit by warm evening sunlight behind.A person sits on a wooden garden bench in a fenced backyard with a tabby cat resting on their lap while holding a red ice lolly. The person looks down toward the cat, which faces the camera. A lawn, shrubs, and a large evergreen tree frame the scene, with warm evening sunlight illuminating the wooden fence and shed in the background.

As someone who engages in a lot of evidence synthesis, and also generally holds a pretty low opinion of the quality of research in the field, this quote upon my re-reading of Meehl and Faust's papers proposing cliometric approaches to metascience struck a chord... 1/6

Screenshot of a page from https://meehl.umn.edu/sites/meehl.umn.edu/files/files/179faustmeehl2002.pdf discussing Einhorn's (1972) study of radiologists' judgments of biopsy slides. The highlighted passage states that although the pathologists generated useful data, they did not make proper use of it themselves. Statistical analysis identified a subset of rated variables that predicted survival time with modest accuracy, whereas the radiologists' overall severity ratings did not predict outcomes. The passage argues that the experts struggled to identify which observations were predictive and how to combine them effectively, illustrating the value of statistical models over unaided expert judgment.

If anyone wants to hear me talk very fast about appraisal of competing theories, riskiness, corroboration, observational uncertainty and error propagation, and meta-analysis for tests of theory predictions... 1/5

If anyone wants to hear me talk very fast about appraisal of competing theories, riskiness, corroboration, observational uncertainty and error propagation, and meta-analysis for tests of theory predictions... 1/5

"The reason most work is bad is that most of us are stupid and incompetent." Man, I feel this in my bones whenever I look around... and it's not even just the usual selection of metascience bug bears that are the reason. It's just that so much is pointless unthinking academic/research busywork.

@lastpositivist.bsky.social · 2mo ago

Since life is keeping me busy I don't have time to do much writing and what not now, so just to keep my blog active I fired off a quick post about the latest report on the humanities. sootyempiric.blogspot.com/2026/06/the-...

With Deborah Mayo, Kent Staley, and Wendy Parker we are editing a special issue of Synthese on "Severity and Learning from Error": link.springer.com/collections/... Deadline for submissions is the 15th of December, 2026. If this is a topic that you are interested in, consider contributing!

Severity and Learning from Error

This Topical Collection examines how inquiry learns from error by focusing on a basic principle of evidence in science, statistics, medicine, law, ...

link.springer.com

For a variety of reasons over the past couple of years the finalisation of the research methods textbook that Eimear Dolan and I have been editing has been somewhat delayed (mostly my fault 😅), but we are back on track and nearing the finish line! 1/8

Well... this tops any peer reviewed publication or other career accolade acheived so far. We got home from an amazing Disneyland trip with the family for @ejps2013 40th birthday to find both a very excited to see us Kiki, and this in the post.

Cover of The Cat magazine, Summer 2026 edition from Cats Protection. A white-and-tabby kitten stands on a log against a blurred green background. Cover lines include “Protect every cat: Why neutering is so important,” “Behind the scenes at a cat centre,” and “In conversation with Bob Mortimer.”Magazine article titled “Celebrating our incredible people” in Cats Protection magazine. The page features a story about volunteer fundraiser James Steele, who completed 495 miles for the Big Winter Roam challenge and raised more than £4,250. A large photo shows him standing beside a stone summit marker on a snowy mountain, and a smaller photo shows him with an orange tabby cat.

Well... this tops any peer reviewed publication or other career accolade acheived so far. We got home from an amazing Disneyland trip with the family for @ejps2013 40th birthday to find both a very excited to see us Kiki, and this in the post.

Cover of The Cat magazine, Summer 2026 edition from Cats Protection. A white-and-tabby kitten stands on a log against a blurred green background. Cover lines include “Protect every cat: Why neutering is so important,” “Behind the scenes at a cat centre,” and “In conversation with Bob Mortimer.”Magazine article titled “Celebrating our incredible people” in Cats Protection magazine. The page features a story about volunteer fundraiser James Steele, who completed 495 miles for the Big Winter Roam challenge and raised more than £4,250. A large photo shows him standing beside a stone summit marker on a snowy mountain, and a smaller photo shows him with an orange tabby cat.

Honestly, might have to just start using github directly for pre-registration as OSF just seems to continually crash on me lately. Every project is in a github repo nowadays anyway and I only link it to an OSF project to utilise the registration.

Trying something out here I'd be interested in peoples thoughts on... Marrying Meehl's distinction of the theoreticians/statisticians problem with Dubin's ontologically agnostic framing of theoretical hypotheses. An idea for explaining the difference between theoretical and statistical hypotheses.

Bild

Trying something out here I'd be interested in peoples thoughts on... Marrying Meehl's distinction of the theoreticians/statisticians problem with Dubin's ontologically agnostic framing of theoretical hypotheses. An idea for explaining the difference between theoretical and statistical hypotheses.

Bild

📣 New preprint! Our latest Project Discover conducted in collaboration with the fantastic team at Discover Strength with delivery led by David Gschneidner and Luke Carlson and of course with the inimitable James Fisher PhD. 1/6

That feeling when the time investment to try to automate data wrangling and cleaning (even with some assistance from an LLM) isn't worth it relative to just doing the wrangling bit manually because the excel spreadsheet you've been given is so ghastly it hurts your soul 🙃

jfc... working through a paper that did performed a quantitative synthesis (badly) at the moment attempting to re-extract data from the underlying studies and it's honestly a mess. I have no idea where the authors got some of their values from... methods are utterly unclear.