Getting Bossy about Jigsaws; “Don’t Fence Us In”

Back in February, I wrote about the “Jigsaw method” of teaching. In this strategy, teachers break a large cognitive topic (say, “the digestive system”) down into small pieces, and assign each piece to a student group.

A closeup view looking along a post-an-rail fence on a prairie of brown and green

Those groups become experts in their pieces — the stomach, the pancreas, the large intestine — and then teach other students in the class about their pieces. When each individual student assembles those pieces into a whole, they have completed the jigsaw; that is, they have understood the full topic.

As I wrote at the time, it’s easy to see potential pitfalls and potential benefits to this method.

When I looked for research on the topic, I found…not much clear guidance either way.

A friend recommended a meta-analysis boasting a HUGE effect size (cohen’s d = 1.20). This meta-analysis, however…

… didn’t include many studies (5, plus 6 student dissertations),

… didn’t appear in a journal that focuses on psychology or education, and

… wasn’t available online.

I’m pretty stubborn, so I kept looking.

The best study I could find, from 2022, found that the jigsaw method provided no benefit — and also caused no harm.

So my conclusion was: “we don’t have conclusive research pointing either way, so I don’t have a strong opinion.”

Since I wrote that post, I’ve gotten some pushback from colleagues I respect — colleagues who, for a number of reasons, think quite highly of the jigsaw method.

Prompted by their concerns, I’ve gone back over this question and made quite a discovery: the pro-jigsaw meta-analysis IS available online. You can find it here.

So, perhaps it’s time to rethink my opinion from February.

But First, a Side Plot

Before I explain my new thoughts about this meta-analysis, I want to explain a few core perspectives that I bring to my consulting work, and to this blog.

In the first place, I’m a very independent person.

I happily seek out new perspectives and new ideas; at the same time, I want to make my own decisions on what to do with those new teaching ideas.

In brief: don’t fence me in.

For this reason, in the second place, I’m also very respectful of other people’s independence.

That is: I don’t want some rando on the internet telling me what to do; and, I don’t want to BE a rando on the internet telling YOU what to do.

Most of my posts include caveats about boundary conditions: “this might have worked in particular classroom circumstances, but they might not work in yours.”

I hope readers find my perspective worth crediting, but you know your curriculum and your students and your school and your culture better than I do. I’m just not going to get bossy without a very persuasive research pool to draw on.

In brief, I don’t want to say “you should do this thing” (or, “you shouldn’t do this thing”) unless I’ve checked out lots of research and found it very persuasive.

I’m not here to fence you in, either.

Back To Our Narrative

So, now that you know my standards, I can explain why finding the meta-analysis has not changed my mind: I can’t read it.

And — this will not surprise you, I hope — I’m not willing to tell you what to do based on a meta-analysis I can’t read.

Here’s the reason for my failure: the meta-analysis is in Turkish. Check out that link above.

Now my point here is easy to misunderstand, so I want to be clear:

I am NOT saying that journals shouldn’t be published in Turkish.

I am NOT saying that research published in Turkish journals doesn’t merit attention.

I AM saying: I can’t read it. And I’m not going to boss you around based on a meta-analysis that I can’t read.

If I could read it, I would have very specific questions: for instance, how on earth did they come up with a Cohen’s d of 1.2? A number that high is almost unheard of.

In fact, a stats-wise friend tells me that–for most psychology topics–a d-value of greater than 1.00 means either a) very small studies, b) bad inclusion criteria, or c) correlation (not causation).

Of course, I don’t know that this meta-analysis includes such concerns. And I don’t know that it doesn’t. My only strong opinion about this study is: people who don’t read Turkish (that’s me!) shouldn’t base opinions on it.

If there’s a reliable English translation floating around, I might revise my thoughts again…


I don’t think we have a clear enough research picture to advocate for or against the jigsaw method.

I suspect it takes an enormous amount of work to get right: so many opportunities for working memory overload! so many chances for distraction!

But, if it’s working for you in your context, the absence of research support should not get in your way. No fences on this horizon…


After I wrote the blog post above, a JUST PUBLISHED study appeared in my news feed. I haven’t reviewed it carefully yet (the full text isn’t available), but here are the authors’ three highlights:

“The Jigsaw method has no effect on students’ autonomous motivation trajectories.

The Jigsaw method does not impact students’ self-regulation over two years.

Collaborative methods are less favorable for students with lower prior achievement.”

This study took place in very specific circumstances — French vocational high schools — and focuses on motivation and self-regulation more than, say, learning. For these reasons, not everyone will find it on-point or persuasive. At the same time, it does include almost 4700 students!

I’m still not going to get bossy (although I confess my doubts and concerns are heightened). But I think this study, along with the 2022 study I’ve already summarized, makes it hard to insist that teachers really must jigsaw right now.


Batdı, V. (2014). Jigsaw tekniğinin öğrencilerin akademik başarilarina etkisinin meta-analiz yöntemiyle incelenmesi. EKEV Akademi Dergisi, (58), 699-714.

Riant, M., de Place, A. L., Bressoux, P., Batruch, A., Bouet, M., Bressan, M., … & Pansu, P. (2024). Does the Jigsaw method improve motivation and self-regulation in vocational high schools?. Contemporary Educational Psychology, 102278.

tags: category: L&B Blog

Leave a Reply

Your email address will not be published. Required fields are marked *