26 Comments
User's avatar
Angela Rubenstein's avatar

These thoughts are so helpful and match my experience 100%! I know you are subbing so have some limitations on setting up your space, but a document camera is the best tool ever for a math class. Of course you have to run around scanning all the desks instead of pulling up the teacher view, but I find grabbing and zooming in on a student’s paper it a better way to set up student thinking as the focus than any snapshot or whiteboard.

Cathy Kelly's avatar

So glad you mentioned the influence of stronger social ties on learning.

I have used Desmos activities and traditional paper based tasks in the classroom. I would observe that in 18 years of teaching my most successful classes were those where there was well developed social cooperative learning... regardless of the platform.

I think this aspect of learning is under researched in the profession.

Michael L. Chrzan's avatar

Another benefit of the group discussion on top of Sepe's points is that it makes it easier for that across-the-room comparison you liked in the digital realm. I'd have even aggregated - start solo, compare and coalesce into one group graph, compare again across the room. In my best planning I'd have even had space for students to note all three versions of the graph (solo, group, class) and then reflect on what they learned in the comparisons.

Also, absolutely fantastic AI points; I appreciate that so much and will be saving this post for future conversations I'm in to reference it. I don't know that I really see any student-level benefits of AI, everything I see ML/AI being useful and effective for is at the teacher-level and up, like helping the teacher orchestrate a rich class discussion like this one.

Steven's avatar

I've been following Dan's work for over a decade. I'm a big fan! And I really appreciate this lesson analysis.

Once upon a time when I taught middle school and my tech tools varied based on the lesson. Big screen always (after SMART boards got phased out). For a "Graphing Stories" lesson, it would be paper for the students and my trusty document camera for me to show off the amazing thinking that I saw happening in the room. Once I had spied out three or four examples of amazing student thinking (common misconceptions, varied levels of precision, etc.), the doc cam made it very efficient for me to lead a compare-and-contrast discussion at the front of the room with the whole class.

These days I am teaching high school and my go-to "tech" tools have shifted primarily to big vertical whiteboards and paper inside sheet protectors (so, not tech). Throughout my read of this lesson study, I found myself thinking, "How would this activity go with trios at a whiteboard working with the paper coordinate grid handout inside a sheet protector?" There is so much about this set up that I love (though it requires a lot of community building and establishing protocols). I could probably answer my own question in the Fall when I return to my classroom.

Anna Blinstein's avatar

As Angela mentioned, there are digital teacher-facing tools that could capture most of the benefits of tech without putting screens in front of students. I'm partial to the iPad + Apple TV combo - affordable, very easy to snap photos of student work and throw several photos into Notability and then project, annotate, and compare on a projector. Added bonus is that the annotated photos can be easily saved to a communal Drive folder as class notes for students.

And as Dan mentioned, having students do this activity in groups on large whiteboards solves some of the tech-free issues in a different (better?) way, although then you lose the ability to see each individual student's thinking.

Different tools for different use cases! Although I do love this activity in its digital form, so many of my students benefit from a digital detox right now.

Dan Meyer's avatar

> I'm partial to the iPad + Apple TV combo - affordable, very easy to snap photos of student work and throw several photos into Notability and then project, annotate, and compare on a projector.

I love that this works for you. I rarely feel clumsy with tech but I admit I have never quite gotten the hang of this particular dance. I need more practice.

Dan Anderson's avatar

Not for nothing but the old school nomenclature of VNPS and VRG (Building Thinking Classrooms for most) would go a long way to improving this non-digital activity over paper and maybe even the electronic version. (Even the random nature of the grouping could be ignored…)

Dan Meyer's avatar

Yeah, interesting. I have been working in a BTC class and hadn't considered VNPS for this particular task. Would you tape a blank graph to the whiteboard? I wouldn't want kids to have to reconstruct the graph and scale the axes by hand here. PS. Dig your recent blogs on AI. I had thoughts but you don't do comments (????) so I had to tell them to my kids.

Dan Anderson's avatar

nah, just have them do their best sketch without a grid. Generally they're not the worst at a decent sketch without a grid... :-)

Thanks and yea no comments on this new blogging platform (grav). Boo. But also it's all static files, so no easy hacking like wordpress. Yay!

Cara Lokken-Frandsen's avatar

BTC is da' bomb diggity!

Julie Pottinger's avatar

As a longtime “Desmos” user (teaching in Steelers’ country, I tell students that much like every true Pittsburgh fan will never call Heinz field “Acrisure Stadium,” I’m not sure that I will ever be able to call Demos, “Amplify”) I appreciate your evaluation of wins vs losses regarding the use of tech. I find myself using Desmos activities less frequently because the use of tech involves so many lost minutes and so much distraction, that the pace of the lesson is lost. However, where Desmos really shines is in creating experiences that are clunky, cumbersome, inefficient to set up, or otherwise distracting in their own right when done offline. One of my all time favorites is click battle. Some of my favorites from this year were the 7th grade lesson on the triangle inequality theorem (the experience of literally seeing the sides fail to meet up stuck with my students all year), the sampling activity “crab island”, and turtle time trials (probably students’ all time favorite). As with anything, we need to be judicious with our use of tech and I hope that Desmos/ Amplify will capitalize on their true strength of providing experiences that are hard to replicate off-screen. One final note. When you talked about sharing in anonymous mode, I thought about the fact that I have all but stopped using it because of the added distraction it creates. While I initially thought it was interesting because it may spark students in learning about a variety of mathematicians, I have now realized that it is doing just that- at exactly the wrong time. Sharing work in anonymous mode is intended to focus students’ attention on the mathematical concepts for the lesson. Instead students are immediately focused on the names- some even googling the names instead of engaging in the intended task. Additionally, the reading difficulty of the names makes it difficult for students to find their names when I project the results of card sorts/ etc. My recommendation would be to provide an optional mode in which students are anonymized as “Student 1” etc. This would eliminate the distraction element, create ease when students try to find themselves in projected work and, as a teacher, it would make students easier to track when I have anonymous mode turned on.

Dan Meyer's avatar

Thanks for these reflections, Julie. I kicked them into a product channel over here and they got a lot of valuable conversation going.

Cara Lokken-Frandsen's avatar

Desmos is a powerful tool and allows kids to demonstrate multiple approaches to the problem solving that the educator can then share with the whole group.

Tommy Joseph's avatar

> Overall, this experience makes me want new ways to mix media, to add a student’s print work to our digital space, for example.

I've been thinking about this as well. iPads are great but they are still screens and they're pricey. I really like the idea of giving each student their own "digital pen". Whatever they write with this pen will be recorded to a digital notebook that the teacher can see on some website. The teacher would have visibility into the student work - without the costs of the laptop transition and without the cultural baggage of the screen.

The [livescribe livepen](https://us.livescribe.com/products/livepen) seems to be the most promising. It's a pen with a camera on it that requires special paper. It costs $80 and you can print the special paper with a laser printer (but must be 300+ DPI). The nuwa pen is also interesting. It doesn't require special paper, but it's more expensive at $350, and there are no anecdotes saying it works.

I've been considering trying to prototype different variants of a "digital pen" myself. One idea is a pen/pencil harness at the top of the pen that has a camera and a button. The student flips the pen, faces it at their work, and takes a picture. Software manages creating/updating their work each time they do this. It's more friction but lower cost (and keeps the smartphones away).

wess trabelsi's avatar

The sandwich-worthy study is incredible data to convince anyone that addressing student use of AI is a total "code blue" situation.

I actually emailed the authors a couple days ago because, as elegant as their data is, the paper leaves a massive blind spot regarding the actual distribution of adoption. Because they relied entirely on de-identified, cell-level log data rather than student-level longitudinal tracking, we don't actually know who is driving this 25% retention drop.

I asked them because I wish the paper could clarify two major unknowns:

1. Widespread Diffusion vs. Outlier Concentration: Does that massive aggregate drop in study time mean every student is leaning on AI a little bit (widespread cultural diffusion), or is it driven by a smaller group of heavy users completely outsourcing their entire workload?

2. Real-World Alignment: How do these behavioral traces actually line up against external, real-world adoption curves for ChatGPT among these age brackets?

whatever the answers are, they would not refute the problem, but would be extremely valuable in designing a strategy to address it. I haven't heard back yet...

Dan Meyer's avatar

Oh interesting. So in the worst case scenario here (empirically speaking) a small number of students could have accounted for a majority of the change pre and post. I appreciate your close eye on research in our field.

wess trabelsi's avatar

See that’s the thing, I’m not sure I would have deemed that the “worst case scenario.” To me that’s just definitely a different problem. Why would that configuration be worse than all the kids using it a bit, or equally? Are they? Why or why not?

You apparently figured this out already, but you may have to spell it out for the rest of us.

Josh Watson's avatar

I think some of the drawbacks of paper could be solved with the right technology, and it wouldn't be horribly difficult to implement. For example, the extra login time could be fixed by biometric login + passkeys + the ability for the teacher to push out a particular website to all students screens at the same time. The metal in the way could be fixed with screens that are horizontally situated (ie, tablet screen, 2-in-1 foldable display, or desk embedded screen). If the difference between writing on paper and writing on the device was minimized (ie, a program that allows active digitizer input), we could get some of the separation distance back.

The real issue now, I think, is that we can do all of these things, but not in a seamless, completely effortless way like handing out a sheet of paper. If these kinds of fixes could be implemented (and others), I think there could be a lot of benefit for students and teachers. I'm still going to walk around the room and interact with students, but a quick note from a trained AI bot that students in group A could use a quick teacher talk because they are stuck trying to graph and group B is already finished and needs an extension might be interesting to explore.

Allison Krasnow's avatar

Great post. We have been talking a lot at work about how rare it is for teachers to teach students HOW to use AI in productive ways. An informal survey of my own 2 kids (middle and high school) supports that. No one has taught them how to use AI. So the only ways they know to use it for school is for cheating which they learn and share new ways among themselves in the absence of any other possibilities. LOVED the Coach Brown clip. I have read many sports coaches biographies and find their wisdom to be spot-on for classroom relationship-building pedagogy. I'm hanging onto that clip to use in professional learning on math leadership. That's for that gem.

Dan Meyer's avatar

> No one has taught them how to use AI. So the only ways they know to use it for school is for cheating which they learn and share new ways among themselves in the absence of any other possibilities.

A lot of people make this argument that teaching students how to use AI in productive ways will discourage cheating, and I just have this ... feeling ... that kids may use AI in productive ways and still cheat. It's just so good at cheating and many kids really want the grades that cheating with AI can produce.

Allison Krasnow's avatar

100%. But I would argue that as adults we probably all to a bit of both as well. Kids are prone to finding the easy way out. I suspect that has been true since the beginning of time. Plenty of adults are too. I've been out of the classroom for more than a minute now, but I suspect that a common teacher response to being asked to teach students to productively use AI would be that they don't have time. Or that they will cheat anyway, or wondering what you'll take off my plate now that teaching them how to use AI is now on my plate. Or simply not feeling comfortable enough themselves using AI to enhance not curtail learning to be confident teaching it to their students. I am empathetic to all of this, nor do I have a perfect solution to propose. But what I do know is that the more opportunities I have had recently to see incredibly productive uses of AI to deepen my own understanding and analytical process, the more fascinated I am of how we can work to find time to explicitly teach these same skills to students with an eye for teaching them when using it is appropriate and when it isn't. I have learned a lot recently from Ed Campos who has done some really great thinking around this in the math space.

Clara's avatar

Highlighting- not blighting!

Clara's avatar

This is exactly what happened in my classroom as we returned after the pandemic. Students did not have their computers, everyone wasn’t looking at the same screen, no one could complete the online activities. It was such a blessing to return to paper! Our school provided white boards for placing around the room so the kids could show their work and everyone could see- a bit clumsy, but it worked for comparing side by side, and for blighting the interesting results so kids could talk to each other about their thinking. The next year we got chromebooks. Yuk. The sweet spot is probably somewhere in the middle- huge chunks of white paper and fat colored markers!

Ji Hoon Chun's avatar

I think augmented reality glasses have the potential to combine the tactile nature of paper with the organizational benefits of screens and digital documents. It would resemble an AR version of Livescribe with more flexibility: the stylus won't need to make physical marks on paper or other surfaces because the glasses will overlay virtual marks.

But obviously there are massive privacy issues if the glasses have cameras.

Cara Lokken-Frandsen's avatar

Hello Dan!

Good on you to do summer school! What lucky kids!

I have found that parents and professionals are moving the pendulum back when it comes to educational technology. The new flex seems to be having a classroom that uses only paper pencil, and many parents appear to prefer that approach as well. Recently a new math teacher was eager to tell me that she only uses paper pencil tasks.

Last fall, I had an impromptu parent-teacher conference in my front yard with a parent who strongly urged me to stop assigning tech-based math homework. I was surprised, but I also understood the concern. If students are not being monitored, ed-tech may not be as effective as intended. Especially now that AI makes it easy for students to bypass the thinking process.

Finding balance between technology and authentic student thinking is challenging. Not long ago, teachers were expected to integrate technology into lessons and teach students how to use it appropriately. Now, there is growing pressure to reduce it.

How do we move forward? I hope we can help parents see that we know they are concerned and that excessive screen time is a problem but the benefits of well monitored digital learning is has advantages when used in balance with paper pencil or white boards and markers. Kids shouldn't have to wait to see their thinking errors for the days it may take for that paper pencil task to be returned when the thinking error can be corrected in the same moment with answer checks. Kids also need to be taught how to appropriately use the digital tools and not cheat their own learning and thinking away. They need to be taught how to work collaboratively with technology. They need to have experience with technology to use it responsibly in their future.

Honestly, seeing this article made me feel better about how to approach this challenge next year. Balance is the key.

Thank you Dan!

I'd love to tell you about the no tech activities we did this summer too. We collected measurement data. The kids launched rubber chickens with therapy bands as far as they could and then we found the mean, median, mode and range of those data sets! FUN STUFF! We also did break out activities where the students solved problems and entered the three or four digit codes into pad locks to solve a mystery. (I also used invisible ink and black lights for some of the clues.)

Dan Meyer's avatar

Rubber chicken modeling! Huge fan!