Our article 'The role of knowledge-rich curricula in promoting deep thinking and complex skill acquisition' is online and free. #EduSky @danielmuijs.bsky.social @timsurma.bsky.social @dylanwiliam.bsky.social www.sciencedirect.com/science/arti...

AI influencers such as Ethan Mollick are already trumpeting the results of a poorly designed World Bank pilot study that employed AI with Nigerian students. This thread provides a fair and measured critique that's worth reading in full. Again I must ask: Why so many shady studies?

Join the Bell Ringer Book Club about our book 'Developing Curriculum for Deep Thinking: The Knowledge Revival'. @timsurma.bsky.social @dylanwiliam.bsky.social #EduSky open.substack.com/pub/hollykor...

In this top resource from March, John Hattie, @dylanwiliam.bsky.social, and Arran Hamilton show how de-implementation can cut ineffective initiatives and focus on what works. Learn how to improve outcomes while restoring balance in this free, on-demand webinar: www.corwin.com/free-...

We zijn nog altijd erg geïnspireerd door het werk van @dylanwiliam.bsky.social ! Samen met @dmasluijsmans.bsky.social vertaalde ik een overzicht van zijn vijf kernstrategieën van formatief handelen. Gratis te downloaden, en heel praktisch 👇 toetsrevolutie.nl/de-vijf-kern...

At researchED Scandinavia on March 15th, Kennet Fröjd hosted a panel discussion on #ArtificialIntelligence and education with Nidhi Sachdeva, Arnold Pears, Eva Hartell and me. You can listen to it on Spotify: lnkd.in/dg8iU6Ht and see it on YouTube: youtu.be/Iq9God1rjOs

NAEP Law uses the phrase "valid and reliable" 8 times and references "professional testing standards" 7 times. We can't measure educational progress with a bendable yardstick. Let's get the expertise we need. www.nagb.gov/naep/the-nae...

Quantitative evidence of the impact of teacher-student relationships on achievement is thin on the ground, so this paper on the positive benefits of "looping" (teaching the same class for more than one year) by matthewakraft.bsky.social and colleagues is important: bit.ly/4bBiBvD

Share widely! @educhatter.bsky.social @rastokke.bsky.social @carlhendrick.substack.com @ehanford.bsky.social @j-metsala.bsky.social @mrzachg.bsky.social @steplab.bsky.social @tombennett71.bsky.social @jim-hewitt.bsky.social @mrsball.bsky.social @dylanwiliam.bsky.social

Looking forward to learning a lot at the @learningandtheb conference in New York City next month: bit.ly/4ixCpSP

Check out this amazing list of speakers at researchED Haninge in March: bit.ly/42F6hbk

15 years ago I did the IES cluster-randomized trials training program at Northwestern--they're still offering it in 2025, and I highly encourage early-career researchers to join! steppcenter.northwestern.edu/education-tr...

Last night I had the pleasure of participating in a roundtable discussion about biologically primary and secondary knowledge and cognitive load theory, with David Geary, katejonesteach.bsky.social, and John Sweller, organized by Kennet Fröjd. You can hear it here: spoti.fi/3Wt8kvc

A difference-in-difference analysis exploiting differences in the timing of charter school entry finds that a 10% increase in charter market share reduced the supply of traditionally-prepared teachers by 1% each year: bit.ly/3E1GPme ($)

Great to see that the scale-up of Embedding Formative Assessment seems to be making a difference in schools in England. The Behavioural Insights Team evaluation found that close to 95% of teachers surveyed were using strategies emphasized in the programme: bit.ly/3WIdwM7. Full report: bit.ly/3CcVlqM

Full disclosure: I am on the project's Advisory Board in a voluntary capacity, and have no financial interest in Aristotal. I just think it's pretty amazing.

For a while now, I have been convinced that one of the most effective uses of AI in education would be to provide personal, private, coaching for teachers. I thought it would take five years, but it seems that it is here, or at least very close, already: bit.ly/3PADiO8.

It’s here! Our new open access book on curriculum written by Tim Surma, @paulkirschner.bsky.social, @dylanwiliam.bsky.social, @timsurma.bsky.social, Nuno Crato, Elizabeth Rata, John Hattie, Claudio Van Hees, Michiel Wils, Jasper Nijlunsing and me. Download it here: link.springer.com/book/10.1007...

Going through my Downloads folder I found a study by Babu and Mendro from 2003's AERA conference. They found that low achievers benefitted greatly from having more effective teachers for three years, but were less likely to get them:

About 8 years ago, @siobhanleahy50 and I prepared a Quick Reference Guide for Embedding Formative Assessment. It's now out of print, so the copyright has reverted to us, and we are free to share it with anyone who wants it: bit.ly/EFAQRG (pdf).

More and more hotels in which I stay have refrigerators that charge you if you move an item. They usually offer an additional refrigerator if you want to keep something cold, so I have decided to ask for a refrigerator every time. If we all do this, they may change their stupid policies...

"If we had an extra week, we could all read Morgan Polikoff’s (2021) Beyond Standards to try to understand the problems caused in our decentralized system by the lack of high-quality curricula and what states might do to evaluate and recommend curriculum materials."

"AI-generated poetry is indistinguishable from human-written poetry and is rated more favorably": go.nature.com/3O9RHAa Image

It seems fair to say that most people in Europe find the result of the US presidential election completely incomprehensible. If you want to understand why it happened, as opposed to lamenting the electorate's decision, this piece by AEI's Rick Hess is a good place to start: t.co/ApyBIO6LMR

It's been a while since I posted a 10 minute chat on GenAI and am delighted to get back into it with a juicy one from @dylanwiliam.bsky.social.

When I trained teachers, I wanted all my students to find their own personal style of teaching, but I also wished I had something that would give them a set of basic techniques to get started. In other words, something like @MrZachG's wonderful new book: amzn.to/3ARrzXH

Last week, I talked to @MonashUni's @timbocop for one of his "10 minute chats on Generative AI". You can hear the conversation here: bit.ly/4hJz0Rq

Analysis of 1,657 4th/5th grade lessons in 317 classrooms in 4 districts finds "students’ exposure to mathematical language varies substantially across lessons" and students make more progress in classrooms where teachers use more mathematical language: bit.ly/3AyE8aq

It's here! We're so excited to share our latest podcast with none other than @dylanwiliam.bsky.social. Insightful, thought-provoking and inspirational. We are so grateful for Dylan for giving us his time and wisdom. We hope you enjoy it! susanfraser.podbean.com/e/talking-to...

Researchers at Meta's Fundamental AI Research (FAIR) team have been working on "self-taught evaluators" that allow large language models to create their own training data: bit.ly/3Usc27i

Nice counterbalanced study of 2nd graders learning 3 and 4 times tables comparing recitation and retrieval (with flash cards). Spoiler alert: retrieval is better: bit.ly/3A4qhIN

In 2020-21 Connecticut students enrolled in in-person sessions on ~50% of available days—less in schools serving disadvantaged students—so effects of in-person schooling are larger than previously estimated: bit.ly/3C2PsvR ($). ResearchGate link: bit.ly/4dYADaU

Just tracked down this, from Abraham Lincoln, in 1832:

"Can we trust the transcript?" by Joe Feldman makes important points about the inequitable nature of many school grading policies but unfortunately assumes teacher-assigned grades and test scores measure the same thing, which they don't—and shouldn't: bit.ly/3Yfqvok

A meta-analysis of 28 studies on chatbot-assisted language learning found an average effect of 0.48 sd which is in line with other studies, but there is considerable heterogeneity in the results: bit.ly/4hj6m9B. ResearchGate link: bit.ly/3CdhSTG

Phonological-only reading instruction works, but not as well as print-based approaches, is—unsurprisingly—more effective for phonological awareness than word reading and, again perhaps unsurprisingly, works less well in preschool: bit.ly/4ffN4AJ ($)

Ten years ago, Yiping Li and Glenda Lappan edited a collection on "Mathematics curriculum in school education". I'd forgotten I'd written an introduction to the section "Curriculum and student learning"—I hadn't even added it to my publication list—but you can now read it here: bit.ly/3TviTMG (pdf)

I’m new to BSky so thought I’d introduce our work. In my lab at the University of Cambridge, we study the development of the #adolescent brain, cognition, social behaviour and mental health. You can find our papers, talks, interviews and more here: sites.google.com/site/blakemo...

My first post on Bluesky! Proud to announce our forthcoming open access book coauthored by Claudio Vanhees, Michiel Wils, Jasper Nijlunsing, Nuno Crato, John Hattie, @danielmuijs.bsky.social, Elizabeth Rata, @dylanwiliam.bsky.social, and Paul A. Kirschner.

I asked ChatGPT4o to give me the data from this graph in CSV format. It replied "It appears that the file you uploaded is an image. I cannot directly extract the numerical data (blue crosses) from the image alone." I asked the free version of ClaudeAI to do the same, and it gave me the data I wanted

Coming soon: new publication on curriculum with Tim Surma, @paulkirschner.bsky.social, @dylanwiliam.bsky.social, Nuno Crato, Elizabeth Rata, John Hattie, Claudio Van Hees, Michiel Wils and me. It's going to be good 😉

Giving professional translators access to LLMs substantially raised productivity and "for every 10x increase in model compute, tasks were completed tasks 12.3% quicker, received 0.18 s.d. higher grades, and earned 16.1% more per minute (including bonus payments)": bit.ly/3XjbGRj

We’ve known for a while that ‘brain training’ rarely leads to any transferable benefits in things beyond the trained task… but this new study takes it a step further and finds no neural (as well as behavioural) changes. ‘Brain training’ doesn’t even change the brain! www.nature.com/articles/s41...

#GeorgeBox @dylanwiliam.bsky.social

I asked ChatGPT4o to represent the data in a table in the form of a bar chart. Here are the table and the chart:

Delighted to announce that, after a 20 year absence, I shall be presenting at the 49th annual conference of the International Association for Educational Assessment in Philadelphia next month: lnkd.in/e_XCqT9b

Earlier this month, I talked with Steph Seidel and Jen Rasmussen for their ChalkTalk podcast. You can listen to the conversation here: spotifyanchor-web.app.link/e/6pxGS7EscMb

One of the ways that test anxiety harms student achievement may be by making students less likely to use practice tests in preparing for high-stakes tests: bit.ly/3yHGfYt ($)

An analysis of physician quality has interesting parallels with research on teacher quality. Physician quality matters (in Norway, 1 sd of physician quality reduces 2-year mortality by 12 percentage points), but it is hard to pin down and not correlated with patient ratings: bit.ly/3MaA8z1

Yes, absolutely this. Pretty much every country in Europe (and many beyond) have a teacher recruitment and retention crisis. So it is not down to the particular accountability regime or policies in England as often stated.