Continuous Reinforcement

What is Continuous Reinforcement?

Continuous reinforcement is a schedule in which every occurrence of a specified response is reinforced. It is defined by the response–consequence arrangement, not by a promise that learning will be rapid. Ferster and Skinner distinguish it from extinction and intermittent schedules.

Reinforcement is not punishment

A reinforcer strengthens subsequent responding. Punishment is not a type of reinforcer. “Negative reinforcement” refers to strengthening behavior through removal or avoidance of an aversive condition, as distinguished in Skinner’s account.

Example and limits

Giving a token after each correct response illustrates a continuous schedule if tokens function as reinforcers in that setting. Giving a token only after several responses illustrates a different arrangement.

Continuous Versus Intermittent Reinforcement

An intermittent schedule reinforces only some responses. Ratio schedules depend on response counts; interval schedules depend on time passing before a response can be reinforced. Fixed and variable versions arrange those requirements differently.

Delivering a consequence after every response can be costly or impractical in some settings. But switching schedules changes the learning arrangement; it is not automatically an improvement.

What happens when reinforcement stops?

The partial-reinforcement extinction effect is greater persistence after intermittent reinforcement than after continuous reinforcement when the reinforcer is subsequently withheld. A history in which some responses went unreinforced can therefore matter when consequences stop. This is a comparison of extinction after different learning histories, not a claim that every intermittent schedule teaches better.

Svartdal’s study of persistence during extinction describes the conventional effect in comparisons between groups, but also examines reversed patterns when schedules are compared within the same participant. The comparison and training history matter; the schedule name alone does not predict an exact rate of decline.

In the token example, a teacher might initially give a token after every correct response, later give tokens after some responses, and eventually stop that token arrangement. These are three different stages: initial teaching, a change in reinforcement schedule, and discontinuing the consequence. To understand what happens at the final stage, consider the learner’s experience across the earlier stages and whether other consequences still support the response.