The word doesn’t mean what almost everyone thinks it means.
In ABA, “positive” means something was added — not that it was nice. And nothing counts as reinforcement unless the behaviour actually goes up. Everything else on this page follows from those two sentences.
The four quadrants, without the jargon
Two questions: is something being added or removed, and does the behaviour go up or down? That is the whole grid.
Positive reinforcement
Add something → behaviour increases
Your child hands you a picture card for juice, and gets juice. Asking goes up.
Negative reinforcement
Remove something → behaviour increases
Your child signs “break”, and the demand is paused. Asking for a break goes up — and that is a skill worth teaching.
Positive punishment
Add something → behaviour decreases
A reprimand after a behaviour. Rarely used in responsible modern practice, and never as a first-line plan.
Negative punishment
Remove something → behaviour decreases
Losing access to a toy after throwing it. Common in everyday parenting; used sparingly and with a plan in good programmes.
A well-run programme lives almost entirely in the first two boxes. If you are told a plan relies on the bottom two, that is a conversation to have in detail before you agree to it.
What parents actually ask about reinforcement
Adding something after a behaviour that makes that behaviour more likely to happen again. The word “positive” is arithmetic, not a value judgement — it means something was added, not that it was pleasant.
The definition is circular on purpose, and this is the part people miss: something only counts as a reinforcer if the behaviour actually increases. Praise a child lavishly and watch the behaviour drop, and your praise was not a reinforcer, however well-intentioned. It is measured, not assumed.
Both increase a behaviour. Positive reinforcement adds something; negative reinforcement removes something. A child who asks for a break and gets one is being negatively reinforced for asking — the demand was taken away — and that is a good thing to teach. Negative reinforcement is not punishment, which is the mix-up almost everyone makes. Punishment decreases behaviour; both kinds of reinforcement increase it.
The difference is timing. A bribe comes first and is offered to stop something that is already happening — a sweet handed over mid-tantrum. Reinforcement comes after the behaviour you want, on a plan set in advance. Confusing the two is understandable, and the practical test is simple: if the reward is being negotiated during a difficult moment, that is a bribe, and it teaches the difficult moment.
Usually satiation — a child who has had unlimited access to something all morning has no reason to work for it at eleven. It can also be that the schedule got too thin too fast, or that the effort being asked for went up without the payoff going up. This is why preference assessments get repeated rather than done once, and why a good programme rotates reinforcers.
No, and a programme that has no plan to fade them is badly run. The sequence is deliberate: reinforce nearly every response while a skill is new, then thin the schedule, then move from artificial reinforcers to the natural consequences of the skill itself — a child who learns to ask for juice gets juice, and eventually that is the whole point. Ask your BCBA what the fading plan is. There should be one.
Keep reading
ABA therapy examples
Reinforcement, prompting and chaining in ordinary family moments.
Discrete trial training
Where reinforcement sits inside a structured teaching trial.
What is ABA therapy?
The field this term belongs to, explained without jargon.
Types of autism therapy
What else belongs in the week beside ABA.
Parent coaching
Learning to run these strategies yourself, between sessions.
Questions families ask
What a session looks like, and how hours get decided.