Reinforcement

Reinforcement is any consequence that follows a behaviour and increases how often that behaviour happens in future. That is the whole definition — and the word increases is doing all the work. A consequence is only reinforcement if the behaviour goes up. If it doesn't, whatever you delivered was something else.

This is what makes reinforcement a slippery idea in practice. It isn't a category of nice things you give people. It's a relationship between a behaviour and what follows it, and you can only identify it after the fact, by looking at what the behaviour does next.

Reinforcement and reinforcer are not the same word

Reinforcement is the process: behaviour occurs, a consequence follows, the behaviour becomes more likely. A reinforcer is the particular consequence doing the reinforcing.

The distinction matters because it stops you calling something a reinforcer before you have any evidence. A sticker is not a reinforcer. A sticker is a sticker — it becomes a reinforcer only once you can show that delivering it made the behaviour more frequent. Until then it is a preferred item, which is a claim about what the learner likes, not about what changes their behaviour.

Positive and negative reinforcement

Both increase behaviour. The difference is whether something is added or taken away.

• Positive reinforcement adds something after the behaviour. A student finishes their work and receives specific praise, and work completion goes up.
• Negative reinforcement removes something after the behaviour. A student asks for a break, the demand is briefly withdrawn, and asking for a break goes up.

Here positive and negative mean added and removed, in the arithmetic sense. They carry no judgement about whether the consequence is pleasant or the behaviour desirable. This single piece of terminology causes more confusion than anything else in behaviour analysis, and it's worth being precise about, because the alternative reading — positive means good, negative means bad — leads people to describe telling a child off as "negative reinforcement" when it is neither.

Negative reinforcement is not punishment

Negative reinforcement makes behaviour more likely. Punishment makes behaviour less likely. They are opposites that happen to share an unfortunate adjective.

A useful check: ask what happened to the behaviour afterwards. If it increased, you're looking at reinforcement of some kind. If it decreased, you're looking at punishment. Nothing about how the consequence felt to deliver tells you which one occurred.

How to tell whether something is actually reinforcing

There is only one test, and it is empirical: deliver the consequence contingently and watch what the behaviour does over the following days.

• If the behaviour becomes more frequent, the consequence functioned as a reinforcer for that behaviour, for that learner, under those conditions.
• If nothing changes, it didn't — regardless of how confident you were, or how much the learner appears to enjoy it.
• If the behaviour decreases, you have inadvertently punished it.

All three qualifiers matter. Reinforcers are specific to the person, to the behaviour, and to the moment. Attention from a teacher can reinforce a behaviour on Monday and do nothing on Friday, because the learner has had plenty of attention already — a shift in motivating operation. A baseline taken before you introduce the consequence is what makes any of this visible.

Reinforcement in the classroom

The most common failure isn't choosing a weak reinforcer. It's delivering a perfectly good one at the wrong moment, and reinforcing something you didn't intend.

Handing a distressed student a preferred item to settle them will reliably increase whatever behaviour immediately preceded it. If that behaviour was shouting, shouting is what gets reinforced — the item's popularity is irrelevant to the contingency. This is covered in detail in The iPad is Not a Reinforcer.

A few things that make reinforcement work in practice:

• Match it to the function of the behaviour. A learner escalating to escape a task is not asking for a token; they are asking for the task to stop. Reinforce the appropriate request with the thing they were actually after.
• Deliver it immediately. Reinforcement strengthens whatever the learner just did. A delay of even a few seconds lets an unintended behaviour slip in between.
• Be specific about what earned it. "You put your hand up and waited" tells the learner what to repeat. "Good job" doesn't.
• Thin it deliberately, not accidentally. Continuous reinforcement builds a new skill; intermittent schedules make it durable. Drifting from one to the other by forgetting is how gains disappear.

For how reinforcement fits alongside prevention and response in a whole-class plan, see the guide to classroom behaviour.

Common mistakes

• Calling a preferred item a reinforcer on a behaviour plan before any data support it.
• Using "negative reinforcement" to describe a consequence intended to reduce behaviour.
• Reinforcing the escalated version of a behaviour because that's the version that got noticed.
• Assuming that what reinforced a behaviour last term still works this term.
• Treating a reinforcer as a property of the item rather than of the effect it produces.