Thomas Phelan's method is one of the best known in this subject and the counting half is what everybody remembers. The book is more careful than its reputation.
How the counting works
The behavior starts. You say "that's one" and wait five seconds. If it stops, you move on. If not, "that's two". If it still has not stopped, "that's three, take five", and the child takes a five-minute break.
The rule that makes it work is the one most often dropped: you only count. No explaining, no repeating the request, no visible irritation. Ask once, then count without emotion. Most reports of it failing are reports of counting while arguing, which is just arguing with numbers in it.
The half people forget
Phelan splits the job in two, and says so plainly: stopping unwanted behavior, and starting wanted behavior. Counting is for the first. He is explicit that it does not work for the second, because starting something takes sustained effort over time rather than a decision to stop in the next five seconds, and recommends routines instead.
| Stop behavior | Start behavior | |
|---|---|---|
| Examples | Whining, arguing, teasing, tantrums | Homework, getting dressed, coming first time |
| What works | Counting, unemotionally | Routine, and something that makes the effort worth repeating |
| Timescale | Five seconds | Weeks |
Why you would use both
They are answers to different questions, and a house usually needs both. If your problem is a child who will not stop, counting is a genuinely good tool and this page is not going to talk you out of it. If your problem is a child who will not start, counting has nothing for you and its author says so.
The start half needs a record. Something that makes the effort visible for longer than the moment it happened, that pays more for the difficult things than the easy ones, and that never goes backwards, because a total that can be lost is one a child protects rather than risks. That is the half this site is about.