Ten minutes into the lesson, a student who was doing fine is staring at the ceiling. Nothing happened. Nobody threw anything, nobody said anything, and the work in front of him is the same work he was doing a moment ago. He just drifted.

Look one seat over. There's a decent chance you'll find the reason there.

Researchers now have a name for this: attention contagion. Over the past few years a small group of careful experiments has tested whether one student's focus, or lack of it, actually transfers to the students sitting nearby. The findings are more specific than the seating advice most of us were handed in our first year of teaching, and one of them quietly contradicts a rule that almost everybody repeats.

What Attention Contagion Actually Means

The term comes from a 2021 study by Noah Forrin and colleagues at the University of Waterloo, published in the Journal of Experimental Psychology: Applied. The setup was small and tightly controlled. Pairs of undergraduates watched a recorded lecture together, except one member of each pair was a member of the research team playing a part.1A confederate is a member of the research team who poses as an ordinary participant and behaves according to a script. Using confederates lets researchers control one variable, in this case peer attentiveness, that would be nearly impossible to manipulate on purpose in a real classroom. The cost is realism. Participants knew they were in an experiment, the lectures were prerecorded, and nobody had any history with the person sitting beside them. Real classmates do have history with each other, which could make these effects larger or smaller than what the studies measured. Every study described in this post except the deskmate field experiment used some version of this controlled setup. Half of these actors were trained to look engaged: leaning forward, taking notes, watching the screen. The other half slouched and looked bored.

Students paired with an engaged partner reported paying closer attention, took more notes, and scored better on the quiz afterward. That result on its own isn't shocking. Somebody visibly checked out in your line of sight is distracting, and distraction hurts learning.

The second experiment is the one worth sitting with. This time the actor sat behind the participant, completely out of view. The effect still showed up. Students couldn't see what their partner was doing, but they could hear it, sense it, and pick up on it well enough that their own attention shifted. Whatever is spreading here, it isn't only a matter of catching movement out of the corner of your eye.

The Lecture Hall Study That Sharpened the Picture

In 2024, Forrin's team published a much bigger and much more classroom-like version of the experiment in Scholarship of Teaching and Learning in Psychology. This one is the most useful study for anyone building a seating chart.

They ran four sessions in a real lecture hall. Each session had 45 genuine participants mixed in with 15 trained actors, all watching the same 30-minute lecture video and taking a quiz on it immediately afterward. In half the sessions the actors were attentive. In the other half they were inattentive.

Two design choices make the results unusually clean. First, phones and laptops were banned, so nobody was reacting to a screen. Second, the inattentive actors were coached to be disengaged without being annoying. They slouched, glanced around, and took no notes, but they didn't fidget, yawn, tap pens, or whisper. Afterward, students confirmed on surveys that they hadn't felt distracted by the people around them.2This detail is what makes the study informative rather than obvious. If the inattentive actors had been fidgeting and yawning, any drop in the surrounding students' performance could be explained by ordinary noise and movement, which nobody needed a study to establish. By stripping out the disruptive behaviors and confirming with post-lecture surveys that students did not feel distracted, the researchers could argue that what spread was the inattentive state itself. The sessions ran in January 2020 with 45 participants and 15 confederates each, for a total of 180 student participants.

Then the researchers varied where students sat relative to the actors. Some sat between two of them. Some sat one row behind them. Some sat far away.

Students seated between two inattentive actors rated themselves as less attentive during the lecture, took fewer pages of notes, and scored lower on the content quiz. The people beside them weren't bothering them. They were just checked out, and it traveled.

The Seat Beside You Matters More Than the Seat In Front

Here's the part that should change how you look at your seating chart.

The effect showed up almost entirely in students seated between inattentive peers. Students sitting one row behind them were largely unaffected. Students sitting far away were unaffected, which surprises nobody. But the row-behind result surprised the researchers, and it should interest teachers, because front-to-back position is the thing most of us fuss over when we redo a chart.

If that pattern holds in elementary classrooms, then who a child sits next to is doing more work than what row they're in. Moving a student from the back to the front is a smaller intervention than swapping the two kids on either side of them.

This also has an awkward implication for group tables. In rows, a student has two neighbors. At a cluster of four, everyone is beside or directly across from someone, and at a table of six there are more adjacent relationships than most of us track. Cluster seating is worth having for collaborative work, but it multiplies the number of neighbors each child is exposed to. If you want the longer version of that trade-off, we covered the arrangement research in a separate post on rows, groups, and U-shape seating.

It Isn't Just Distraction

So if the inattentive students weren't making noise, weren't on devices, and weren't reported as distracting, what actually moved?

The researchers looked at this directly. Their analyses pointed toward something closer to social reading of the room. Students seated between disengaged peers rated those peers as less motivated, and there are signs that students partly recalibrate how important the lesson seems based on what everyone around them is doing. Kids glance sideways, see that nobody's writing anything down, and quietly conclude this must not be the part that matters.

A 2025 study by Michael McHale and Michael Kane in Collabra: Psychology found something consistent in an online setting. They had 352 participants watch a virtual lecture alongside on-screen peers who were either engaged or checked out, and interrupted the lecture twelve times to ask what participants were thinking about. Students watching with disengaged peers reported more off-topic thoughts. The contagion doesn't stop at posture. It reaches into what students are actually thinking about while you're talking.

Inattention doesn't just look bad in a room. It tells the other children what the lesson is worth.

The Finding That Contradicts the Advice You've Heard

Almost every teacher has been told some version of this: seat the distractible child next to a focused one. It sounds obvious, and it's been handed down for a very long time.

The 2024 lecture hall study predicted that focus and lack of focus would both spread. It found evidence for only one of them. Inattentiveness was contagious. Attentiveness wasn't, at least not measurably.

I'd be careful about treating that as settled, and so were the researchers.3A null result is not proof that something never happens. The 2024 study could mean attentiveness genuinely doesn't transfer, or that it transfers too weakly to detect at this sample size, or that a room where most people are already paying attention leaves little room left to improve. Earlier work found differences favoring the attentive condition, but those studies compared attentive peers against inattentive ones rather than against a neutral baseline, so the whole difference could be coming from the inattentive side pulling students down. The practical read is that separating disengaged students from each other rests on firmer evidence than seating a distractible student next to a focused one. A single null result is not the same as proof that focus never rubs off. But the direction of the evidence is consistent enough to change what you lean on. Separating students who pull each other off task is well supported. Counting on a focused child to lift a drifting one is a bet with much less behind it.

There's a second reason to be cautious about rescue pairings, and it has nothing to do with the data. Assigning a child to sit next to a struggling classmate is a job. It's usually given to the same handful of kids year after year, and it's rarely presented to them as a choice. In a classroom built around inclusion, that's worth thinking about on its own terms.

What a Semester-Long Seating Experiment Adds

The attention contagion studies all measure a single lecture. Tamás Keller and Felix Elwert ran something with a much longer horizon, published in Sociological Science in 2023.

They randomized deskmates in 195 classrooms across Hungarian schools, covering roughly third through eighth grade, and held those assignments for a full semester. That's 3,365 students, pre-registered, with teachers confirming that 94 percent of students actually sat where they were assigned. As far as classroom seating research goes, this is about as good as the evidence gets.

The average effect was nothing. Sitting next to a higher-achieving deskmate didn't move test scores or grades for the typical student.

The extremes were a different story. The lowest-achieving students did better when seated next to the highest-achieving deskmates. The highest-achieving students did worse when seated next to the lowest-achieving ones. The effects were real, but they were concentrated at the ends of the range, and they ran in both directions.

Put that next to the attention contagion work and a practical picture forms. Neighbor effects exist, they're uneven, and any fixed pairing parks the cost on somebody in particular. Which is a strong argument for moving students around rather than finding one perfect arrangement and freezing it.

The ground under all of this is still thin for younger students.4Almost none of the attention contagion research was done with elementary students. Those experiments used undergraduates watching lectures, and a lecture is a poor match for a second-grade classroom in nearly every way that matters. The deskmate field experiment is the closest thing to a school-aged test, but it measured grades and test scores over a semester rather than attention during a lesson. Younger children have less developed self-regulation, which could make them more susceptible to a neighbor's behavior. Or the shorter, more varied structure of an elementary day could blunt the effect. Nobody has run the study that would tell us, and the researchers themselves have called for exactly that. The largest look at real classrooms comes from a 2026 analysis of PISA data covering more than 230,000 fifteen-year-olds in 70 countries, which found that peer device distraction predicted lower math achievement and weakened the benefit of students managing their own notifications. It's correlational and it's about phones, but it's a signal that the effect isn't just a lab artifact.

What to Do About It on Monday

Fix the neighbors before you fix the layout. If you have limited energy for seating changes, spend it on adjacency rather than rows. Two students who reliably sink each other's focus should not be side by side or directly across a table, and that single change is better supported by evidence than most of the rearranging teachers do.

Separate rather than recruit. Splitting up a pair of students who drift together is the move with real evidence behind it. Seating a distractible child next to your most focused one is a hopeful guess, and it puts a burden on the focused child that nobody asked them to carry.

Rotate on a schedule you'll actually keep. The deskmate research is a reminder that neighbor effects cut both ways. A fixed chart that works well for most of the class is quietly costing whoever ended up in the bad pairing. Changing neighbors every few weeks spreads that around, and it gives you real information about which combinations work.

Count how many neighbors each child has. This is easy to overlook. A student at a table of six is exposed to five people. The same student in rows is exposed to two. If you have a child whose focus depends heavily on who's nearby, the arrangement itself is part of the accommodation, not a separate question.

One thing this research does not support is isolating a struggling student. Pulling a desk away from the group is visible to every child in the room and reads as a punishment, whatever you intend by it. The evidence points toward changing who a student sits beside, not removing them from the seating chart.

Using Inclusive Seating to Manage This

Tracking adjacency by hand gets unmanageable fast. Three or four pairs you're trying to keep apart, a couple of students who need to be near your desk, and a room that isn't a neat grid, and you're doing constraint satisfaction with a pencil at 4pm on a Friday.

That's the specific problem Inclusive Seating exists to solve. In the roster, each student can carry Must not sit near restrictions naming particular classmates. Those restrictions belong to the student, so they hold every time you generate a new chart. You name the pairing once, and it stays broken up for the rest of the year without you having to remember it.

The Randomize algorithm also penalizes repeat neighbors and same-seat placements from the previous chart. Rotation is the recommendation that teachers abandon first, because doing it well by hand is tedious and doing it badly recreates the same clusters you were trying to break up. Here it happens on its own. Click Randomize, and students get genuinely different neighbors while every accommodation you've set stays intact.

For students who need adult proximity rather than peer proximity, the Must sit near restriction can target your teacher's desk directly, along with other elements like the whiteboard, the calm corner, or the library. Given how weak the evidence is for peer focus transferring, anchoring a student to you is a better-grounded strategy than anchoring them to a classmate.

And if you want to experiment with how many neighbors each child has, the magic wand button in the classroom grid cycles through eight preset layouts, from traditional rows to grouped tables to a U-shape. Switching between them takes a click, and your student restrictions carry over to whichever one you land on.

It's free, there's no account, and nothing leaves your browser. Which means you can try three different neighbor configurations on a Tuesday night and keep the one that looks right.