"It's the only thing that works"

Start here, because it's true, and because it's the truth that keeps the habit going. The shout works. If it didn't, you'd have stopped years ago.

There's research that says so plainly. Daniela Owen, Amy Slep and Richard Heyman reviewed forty-one studies of what parents do to get children to comply, with children from eighteen months to eleven years. Reprimands, meaning telling a child off, and negative responses like a stern face consistently produced more compliance. Praise and positive responses had mixed results. So the evidence agrees with you. In the moment, a telling off gets the shoes on more reliably than a "well done" does.

I'm not going to argue with that. I'm going to ask what the shout is working on.

Go back to Tuesday. You asked from the kitchen. Nothing. You asked from the doorway. "In a minute." You asked with the full name. The sigh. Then the shout, and the table. Four instructions, one result. From your side, the shout was the one that worked. Now look at it from the child's side. They received four signals, and three of them, by experience, could be ignored without consequence. The only one that was ever followed by anything was the fourth. A child who's lived through that sequence a few hundred times has learned something precise, and it isn't that they should lay the table. It's that laying the table becomes necessary at the shout and not before.

So the shout works because the ladder has made it the instruction. The first three asks aren't instructions any more. They're the run-up. Your child isn't being defiant when they ignore the first ask. They're responding accurately to what the household has taught them, which is that the first ask is information about your mood and the fourth is the one that counts.

This is my reading, but there's a large body of research behind the general shape of it. Justin Smith and colleagues followed seven hundred and thirty-one families from when the children were two until they were eight, observing how parents and children dealt with each other. Where parent and child fell into what the researchers call coercive interaction, each escalating to get the other to give way, the child's non-compliance grew over time. The coercion amplified the not listening. It didn't reduce it. And the children's own difficult behaviour didn't consistently predict more coercion, which is to say that the pattern was more driven by the exchange than by the child.

So the shout works tonight and makes tomorrow harder. That's not a paradox. It's a training programme, run by accident, in which the reward for ignoring the first three asks is that nothing happens, and the signal for action is a parent losing control. You've been running it for years, and it works exactly as designed.