Field Guide · Glossary
Reinforcement schedule
Learning scienceThe rule that decides which responses earn reinforcement, such as every one, every few, or the first after a period of time.
How it works
Reinforcing every correct response is a continuous schedule, and it is how new behaviors are usually taught. Anything less than every time is intermittent. The intermittent schedules have formal names based on whether they count responses (ratio) or elapsed time (interval), and whether the requirement is fixed or varies. Moving from continuous to intermittent is called schedule thinning. Done too quickly it produces ratio strain: a dog who slows down, pauses or stops working because the payment has become too scarce, which is commonly misread as stubbornness. The usual fix is to return to more frequent reinforcement and then thin more slowly.
What it looks like
A dog learning to sit is paid for every sit. Once the sit is fluent, the handler begins paying roughly every second or third one, and then varies it. If the dog starts sitting slowly or wandering off, the schedule has thinned faster than the dog can follow, and the handler goes back to paying more often for a while.
Commonly confused with
- Rate of reinforcement
- A schedule is the rule for which responses get paid. Rate of reinforcement is how often payment actually arrives over a stretch of training. Two sessions on the same schedule can have very different rates if the dog responds at different speeds.
Related terms