“Positive reinforcement” is used almost everywhere as a synonym for being nice to a child. It is not, and the difference matters because it explains why sometimes nothing seems to work.
A reinforcer is whatever, delivered after a behavior, makes that behavior happen more. It is defined by what follows, not by how pleasant it looks. If you have praised for two weeks and the behavior has not moved, praise is not a reinforcer for this child right now, and continuing does not make it one.
“Positive” here just means something is added after the behavior — a sound, an object, attention. (Negative reinforcement works by removing something instead; both can strengthen behavior, and neither name is a judgment about whether the behavior itself was good.)
Figure — Observe the effect across new opportunities
- Response
- Consequence
- Observe subsequent opportunities
The four things that decide it
What it is. You find out by watching: what does the child choose when they can choose? What do they go back to on their own? Put three things out and see which they take first, several times. That beats any list of rewards — and it tells you what the child prefers, not yet that it works as a reinforcer. The next step, below, is checking whether delivering it actually moves a behavior.
When it arrives. This is where most attempts fail. A reinforcer that lands three seconds late risks strengthening whatever the child happened to be doing in that gap, not the response you meant to reward — it doesn’t always land on the wrong thing, but there is no way to tell which trials it caught. Immediate means in the same moment, not at the end of the session.
What it depends on. It has to follow the behavior and not the passage of time. If it arrives anyway, the child learns that waiting works.
How much there is. A small piece keeps its value; a full bowl destroys it in two minutes. There is no fixed correct portion — it depends on the child and the reinforcer — but a piece small enough to leave room for another trial tends to hold up better than one that ends the interest outright.
Satiation, and its opposite
Nothing exhausts a reinforcer faster than heavy use. If bubbles are the best thing in the world on Monday and nothing by Wednesday, satiation is one possible reason — not the only one. Fatigue, a change in whatever surrounds the activity, or genuinely losing interest can look the same from the outside; watching what happens when you swap in a different reinforcer is what tells them apart.
One simple remedy: keep three or four reinforcers in rotation rather than relying on one. If teaching depends on a single strong reinforcer, using it for everything all day empties it out by the time you actually need it — the fix is variety, not cutting the child off from things they already enjoy.
The reverse also holds: how much a child wants something depends on their current state, and that state changes across the day — a snack offered right after a meal lands differently than the same snack an hour later. That is a reason to notice timing, not a reason to withhold food or access in order to build up want: deprivation as a deliberate technique is a clinical decision, not a home strategy.
What can and cannot be claimed
That a behavior increased under a reinforcer is a checkable fact. That the child “is motivated” or “enjoys it more” is a different claim, and no tally sheet can support it.
One week of watching is a rough check, not proof. Plenty of things can move a count besides the reinforcer — a good day, a change at school, plain variability. What makes the comparison worth trusting is repeating it under otherwise similar conditions: does the behavior rise again when you reintroduce the reinforcer after a break, and drop when you take it away? A single before/after does not rule out coincidence; a pattern that repeats does the real work.
And one limit worth keeping in view: none of this replaces finding out why a difficult behavior happens. Reinforcing an alternative without knowing what maintains the original is a shot in the dark, and that part belongs with your Instructor or a supervising BCBA rather than with home practice.
How it ends
The goal is an activity that sustains itself on what it produces, not a child who works for tokens forever: the request that gets the thing, the game that continues, the task that finishes.
So reinforcement is thinned once the behavior is solid, but only then. Thinning while the skill is still being learned is the commonest mistake in the other direction, and it produces a child who stops trying exactly when they were starting to succeed.
Further reading
- Errorless learning — why you help before the child fails.
- ABA therapy activities — where all of this fits into ten minutes.