By Clay Shumate · World History, Greene County Schools

One voice, fifteen seconds.
One voice, fifteen seconds.

Yelling works.

I want to start there, because pretending it doesn’t is how you lose every teacher who has ever done it — which is every teacher. You raise your voice, the room snaps quiet, and for about eleven seconds you have order. That moment is real. It is also the entire problem.

Because what you just bought with those eleven seconds is expensive, and the bill comes later in the period, later in the week, and later in the year.

What actually happens when a teacher yells?

A reprimand suppresses behavior in the moment without reducing it over time. That’s not an opinion, and it’s been measured directly in classrooms.

Researchers followed 149 teachers across 19 elementary schools in three states, observing 311 students considered at risk for emotional and behavioral disorders. Using a cross-lag analysis — which tests whether one thing predicts another later — they found that teacher reprimands did not decrease students’ future disruptive behavior or increase their engagement, and that while reprimands may briefly suppress misbehavior, they don’t appear effective at reducing disruption or improving engagement over time (Caldarella et al., 2021).

Read that carefully, because it’s a strange result. The reprimand does stop the behavior right now. It just doesn’t produce a calmer room tomorrow. You’re not building anything. You’re renting quiet by the minute, at a rate that goes up.

Why does yelling at one student make the whole room worse?

This is the part that most directly answers the question in the title, and it’s the oldest finding in classroom management.

In the 1950s, psychologists Jacob Kounin and Paul Gump documented something counterintuitive: when a teacher responded to one disruptive student with strict disciplinary measures, that student might stop — but other students would start exhibiting the same misbehavior. They called it the ripple effect, and their conclusion was that a teacher trying to control ripple effects does it better through clear instruction than through pressure on the individual (Kounin & Gump, 1958).

So the volume doesn’t contain the disruption. It broadcasts it.

Everyone in the room just watched what happens when the adult loses composure. Some students file that away as a threat. Some file it away as entertainment. A few file it away as a button they now know how to press. None of them file it away as I should get back to work.

That is what I mean by manufactured chaos. The yelling didn’t respond to disorder. It produced more of it, in more people, a few minutes downstream.

The escalation game

There’s a specific trap waiting for a teacher who tries to win an argument with a student who has decided not to comply.

Some students arrive at school already fluent in coercion — they’ve had years of practice at home learning exactly how hostile exchanges escalate and which behaviors eventually make an adult back down. When a teacher issues an instruction these students don’t want to follow, they escalate their noncompliance to higher and higher intensity until the instruction is withdrawn — what researchers call the behavior escalation game, and it’s a game teachers cannot win and should not play (Walker, Ramsey, & Gresham, 2004; Walker, 1995).

The escalating command — “You will do what I say.” “I told you to start now.” — is the most common version of playing anyway.

Here’s the thing about that game. You can only win it by escalating past a student, and there is no version of an adult escalating past a fifteen-year-old that ends with either of you better off. The student loses standing in front of thirty peers. You lose the room. Somebody’s writing a referral.

The cycle: how it feeds itself

The most important research here comes from parenting, not teaching, and I’m going to be upfront about that limitation rather than pretend the transfer is airtight.

Wang and Kenny followed 976 two-parent families and their children over two years. Harsh verbal discipline by mothers and fathers at age 13 predicted an increase in adolescent conduct problems and depressive symptoms between ages 13 and 14 — and a child effect ran the other direction too, with misconduct at 13 predicting increases in parents’ harsh verbal discipline over the same period (Wang & Kenny, 2014).

Both arrows. The shouting produced more misbehavior, and the misbehavior produced more shouting. That’s not a discipline strategy, that’s a loop with no exit.

Two more findings from that study are worth sitting with. The negative effects of verbal discipline over two years were comparable to those found in studies of physical discipline over the same period. And warmth did not moderate the relationship at all — meaning the belief that it’s fine because they know you care about them does not hold up. The kids who knew they were loved got the same outcomes.

Again: that’s a study of parents, not teachers. But if a caring adult with a lifelong bond can’t yell without cost, the case that a teacher gets a pass is thin.

So why do we keep doing it?

Because it’s reinforced, and the reinforcement schedule is brutal.

The behavior stops. You got quiet. If misbehavior stops, the teacher’s reprimanding behavior is reinforced, and teachers continue to use reprimands (Caldarella et al., 2021). You’ve just been rewarded for the thing that’s making your year harder, and the reward arrived within a second while the cost arrives in three weeks. Nobody’s brain is built to weight that correctly.

There’s also a cost the teacher absorbs directly. Research in classrooms implementing schoolwide positive behavior supports found that teachers with lower ratios of positive to negative interaction, who used higher rates of harsh reprimands and had higher rates of disruption, reported higher levels of emotional exhaustion (Reinke, Herman, & Stormont, 2013).

Which direction that runs is genuinely hard to untangle, and I won’t claim more than the data supports. But if you’ve ever driven home from a day where you yelled and felt hollowed out, that feeling is in the literature.

Would you accept it?

Set the research aside for a second.

Picture your last faculty meeting. Now picture an administrator standing at the front of it and shouting at you, by name, in front of every colleague you work with, for checking your phone.

What would you do? Would you feel corrected? Would you feel motivated? Or would you spend the rest of that meeting composing what you wished you’d said, deciding exactly how much you’re willing to give that person going forward, and telling three people about it in the parking lot?

Now recognize that you’d be doing that as a fully-formed adult with a mortgage and a retirement account and every reason in the world to keep your composure.

We’re asking a fourteen-year-old to handle the same thing better than we would. They’re not going to, and it isn’t because they’re immature. It’s because being humiliated in front of your peers is a real injury at any age, and the response to it is the same at any age. The difference between us and them is not the reaction. It’s only the number of times we’ve practiced swallowing it.

What to do instead

None of this is useful if the alternative is just “don’t yell.” Here’s what actually replaces it.

Go quieter, not louder. Dropping your volume forces the room to work to hear you. It also signals that you are not activated, which is the single most stabilizing thing an adult can communicate in a loud room.

Count out loud, calmly, and let it take as long as it takes. I’ve had it take fifteen seconds. Fifteen seconds feels like a career. It’s fifteen seconds.

Recognize the first person who complies, out loud and specifically. Behavior-specific praise is the most direct substitute for a reprimand, and it works on the whole room instead of one student. Three years of classroom research found a direct positive relationship between teachers’ praise-to-reprimand ratios and students’ on-task behavior, including in middle school and among students at risk for behavioral difficulties (Caldarella et al., 2020; Caldarella et al., 2023). I’ve written about how I use this in classroom management for new teachers. classroom management for new teachers.

Correct privately whenever the situation allows it. A public correction has two audiences and only one of them is the student. (For more on how to handle correction while preserving respect in the classroom, that piece covers the close-range redirect and why it works.) Pull alongside the desk. Say it low. Walk away. You’ve preserved their standing, which means they have nothing to defend and no reason to perform.

Take the exit ramp instead of playing the game. When a student escalates, the move is to give them a way to comply that doesn’t cost them face: a choice between two acceptable options, a delay, a job to do. You are not conceding. You’re declining to play a game you can’t win.

And when you do yell anyway

You will. I have.

The repair matters more than the lapse, and it’s simpler than people think. You go to the student — privately, not as a performance for the room — and you say something close to: I raised my voice at you and that was on me. I’ll do better. Here’s what I still need from you.

That’s it. No speech. No trading your apology for their apology.

I’ll tell you what that does, because it surprised me the first time. It teaches a young person what accountability looks like from someone who has power over them, which for some students in your room is genuinely new information. And it costs you nothing you were actually using. Adults who apologize don’t lose authority. Adults who never do just quietly lose the room instead, one student at a time, in a way nobody announces.

Frequently asked questions

Is raising my voice ever appropriate?
Yes — for safety. A fight, a lab accident, something about to get somebody hurt. Volume is a legitimate emergency tool, and it works in an emergency precisely because you don’t use it the rest of the time. Spend it on Tuesday and it isn’t there on Friday.

What if my students only respond to yelling?
That’s usually a sign the room has been trained on it, not that the students require it. The escalation research explains why: if volume is the only signal that means this time I’m serious, everything below that volume has been taught to mean not yet. Rebuilding takes a few weeks of consistency, and the first two are the hard ones.

Does this mean I shouldn’t have consequences?
No. Calm and permissive aren’t the same thing. A consequence delivered quietly is usually more effective than one delivered loudly, because the student can actually hear it instead of managing their own adrenaline. For a deeper look at consequences that actually teach responsibility, see Teaching Responsibility Without Treating Teenagers Like Children.

I’m an administrator. What should I do about a teacher who yells?
Almost certainly not write them up first. Yelling is usually a symptom of a teacher who’s out of tools and out of gas, and punishing them replicates the exact dynamic the article is about. Ask what’s happening in that room. Then get them a real strategy, coverage to observe someone calmer, or both.