You are committing exactly the same aggregation fallacy that I talk about in the blog post. Your chart shows a relationship between averages!

I'm also moving my personal blog onto Substack so do subscribe if you would like to. profbeckyallen.substack.com

Ha! I’m wildly more optimistic than you about whether AI can re-construct school-level curricula, when asked. It may reflect our subject differences - “English” is very poorly specified. I’m talking about it at researchED- see you there!

We’re going to write a post very soon on confidence interval reporting pros and cons. I like that framing - most framing of CIs is unintelligible to parents.

Thanks to @gatsbyed.bsky.social for funding this update. Full report here - teachertapp.com/uk/publicati.... School Dash blog here - www.schooldash.com/blog-2507.ht...

For heads and students, this is good news as it becomes easier for leaders to fill posts with suitably qualified teachers.