HeyJaredBecky Allen — Falsifiable
On learning science, instructional design, and the future of schooling.
- Indexed issues, last 90 days
- 4
- Latest publication
- Sep 19, 2026
- Audience
- Checking…
- Earliest in this view
- Jul 23, 2026
Latest issues
Where Does the First Lesson Come From? (opens the original)
Read excerpt
You are teaching a new topic for the first time. How do you build your lesson plan? Take something from a textbook or online guide? Use a standard internal lesson plan written by a colleague? Apply a template set of principles you use to teach any new concept in the subject? Whatever you find and adapt, chances are you’ll make different decisions to other teachers.Those of us who are designing instructional platforms face exactly the same question of where we should get the knowledge for that fi
The Theories That Nobody Kills (opens the original)
Read excerpt
I was lucky enough to come of age as an education researcher in the 2000s, as one of those with early access to the National Pupil Database. Cohorts of half a million students gave us the power to detect the impact of almost anything. And, predictably, we found an effect for almost everything!At the time, this didn’t give me pause-for-thought. Nobody had put Meehl (1978) onto a reading list of mine. Our philosophy-of-science training seemed largely detached from the estimates we were busy produc
An Hour on the Volume of a Prism (opens the original)
Read excerpt
Sometimes Craig Barton’s brilliant Mr Barton Maths Podcast drifts into topics that are a little niche, even for me. A full hour of Kris Boulton and Craig discussing how to introduce the volume of a prism? I was fairly sure I had better things to do with my time. Turns out, I was wrong!The backstory to the podcast concerns an early career teacher who is working within Kris’s Engelmann-influence Unstoppable Learning programme. They had produce an introductory sequen
The Teacher Who Cannot Tell (opens the original)
Read excerpt
Last week, I asked a teacher to revise an AI-generated lesson that had been built from a set of context documents. The content looked accurate, the examples made sense and the questions were doing what they were supposed to do. But they changed it anyway, and they justified every edit as they did so. I liked their changes, but I also wondered which of them would actually improve learning.In the previous post I argued that simulated students cannot solve education’s verification problem. An LLM p
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
No verified audience measurement yet.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.