Reinforcement Schedules Made Simple: FR, VR, FI, and VI

Schedules of reinforcement look like four acronyms and turn out to be one two-by-two grid. Once you see the grid, the exam questions in this area become nearly automatic.

The grid

Two questions define every schedule:

1. Is reinforcement delivered after a number of responses (ratio) or after a period of time (interval)?
2. Is that requirement fixed (the same every time) or variable (an average)?

That is the whole system. Everything else is consequence.

Continuous reinforcement (CRF)

Before the four, there is FR-1: every single correct response is reinforced. This is continuous reinforcement, and it is what you use when teaching a brand-new skill.

• Strength: fastest acquisition
• Weakness: rapid satiation, and very fast extinction once reinforcement stops

You do not stay on CRF. You thin the schedule as the skill establishes — that transition is a standard part of skill acquisition programming and a common exam scenario.

Fixed ratio (FR)

Reinforcement after a set number of responses. FR-5 means every fifth correct response.

Response pattern: High, steady rate, followed by a post-reinforcement pause — a brief break right after the reinforcer is delivered before the next run begins. The larger the ratio, the longer the pause.

Session example: A token after every five correct sight words.

Exam cue: "Pauses after earning the reward, then works quickly again" is FR almost every time.

Variable ratio (VR)

Reinforcement after an unpredictable number of responses, averaging out to n. VR-5 means an average of every five responses — sometimes three, sometimes eight.

Response pattern: High, steady, and no meaningful pause. The learner cannot predict which response pays, so the rate stays consistent.

Resistance to extinction: Highest of all four schedules. This is why VR is the schedule slot machines use, and why behaviors accidentally on a VR schedule are so hard to reduce.

Session example: Praise delivered after an average of every four appropriate requests.

Exam cue: "Steady, high rate that persists even when reinforcement stops" is VR.

Fixed interval (FI)

Reinforcement for the first response after a set period of time has elapsed. FI-5 means the first correct response after five minutes.

Note what this does not mean: reinforcement is not automatic at five minutes. The learner must respond after the interval ends.

Response pattern: A scallop — little responding early in the interval, accelerating sharply as the interval nears its end.

Session example: A check-in reinforcer available for the first on-task response after each ten-minute block.

Exam cue: "Little activity at the start, a rush at the end" is FI. Studying only the night before an exam scheduled at a fixed date is the classic human illustration.

Variable interval (VI)

Reinforcement for the first response after an unpredictable amount of time, averaging n. VI-5 means an average of five minutes.

Response pattern: Low to moderate, but very steady. Because the learner cannot predict when the interval ends, there is no scallop.

Session example: A teacher circulating and praising whoever is on task, roughly every few minutes.

Exam cue: "Consistent, moderate responding over a long period" is VI.

Summary table

Two facts carry most of the exam weight here: VR produces the highest, steadiest rates and the greatest resistance to extinction, and FI produces the scallop.

Choosing a schedule in practice

• Teaching something new: CRF. Every correct response reinforced.
• Building fluency once the skill is acquired: thin to FR, then to VR.
• Maintaining a well-established behavior: VR or VI, because they resist extinction and are sustainable for staff.
• Reducing dependence on constant delivery: thin gradually. Jumping from CRF to VR-10 typically causes the behavior to collapse — that is called ratio strain, and the fix is to back the requirement down and thin more slowly.

The RBT boundary applies here as it does everywhere: you implement the schedule specified in the program. You do not decide to thin it because the session is going well. If the schedule seems wrong for the learner, that is a supervision conversation.

Where the accidental schedules live

The most useful real-world insight in this topic: problem behavior is often on a variable schedule nobody designed.

A child who screams and is sometimes given the tablet — not always, just sometimes — is on a variable ratio schedule. That is precisely the arrangement that produces the most persistent, extinction-resistant behavior in the entire grid. It also explains why inconsistent implementation of an extinction procedure is worse than not starting it: intermittent reinforcement during extinction strengthens exactly what you were trying to reduce.

This is why fidelity matters, and why "everyone on the team runs the plan the same way" is a clinical requirement rather than a bureaucratic one.

Quick self-check

1. A learner works quickly, earns a token, pauses, then works quickly again. Which schedule?
2. Which schedule produces the greatest resistance to extinction?
3. A learner shows almost no responding for the first several minutes, then a burst. Which schedule?
4. You are teaching a brand-new skill. Which schedule do you start with?

Answers: FR; VR; FI; CRF. Miss any of them and reread the grid — this is one of the few RBT topics where pure memorization genuinely closes the gap.