Reinforcement schedules control how behavior is built, maintained, and lost. Choose the wrong schedule at the wrong moment and you’ll see slow acquisition, ratio strain, or a behavior that extinguishes the moment you thin delivery. This guide covers the five core schedules, a selection framework for each phase of learning, and the thinning mistakes that silently undermine otherwise well-designed programs.
The Five Core Reinforcement Schedules
Schedules fall into two categories: continuous reinforcement and the four intermittent types.
Continuous Reinforcement (CRF)
Every instance of the target behavior produces reinforcement. CRF is the standard starting point for acquisition. When you’re teaching a new behavior, the learner needs to establish the contingency clearly before you add any ratio or interval requirement. Using an intermittent schedule before the behavior is solid slows learning and introduces unnecessary frustration.
Use CRF when:
- Teaching a new skill (early DTT trials, initial mand training)
- Re-establishing a behavior after extinction
- The learner’s responding is inconsistent or the contingency isn’t yet clear
Fixed Ratio (FR)
Reinforcement is delivered after a fixed number of responses. FR-2 delivers reinforcement after every second response; FR-5 after every fifth.
FR schedules produce high, steady response rates with a characteristic post-reinforcement pause: the learner slows down immediately after receiving reinforcement, then accelerates as they approach the next ratio requirement. The pause is a schedule artifact, not a behavior problem.
Use FR when:
- Thinning from CRF toward intermittent delivery
- Tasks requiring consistent output (matching, academic responding)
- Rate building is an explicit goal
Variable Ratio (VR)
Reinforcement is delivered after an unpredictable number of responses that averages to a set value. VR-3 means reinforcement arrives, on average, every three responses — but it might come after one response, or after five.
VR schedules produce the highest, most stable response rates of any schedule and are notably resistant to extinction. Slot machines run on VR schedules for exactly this reason.
Use VR when:
- Maintaining established behaviors over time
- Natural language and social behaviors where responding should be fluent and persistent
- Extinction resistance is a clinical goal
Fixed Interval (FI)
Reinforcement is available after a fixed amount of time elapses, then delivered for the first response. FI schedules produce a “scallop” pattern: slow responding early in the interval, then acceleration as the interval end approaches.
FI schedules are relatively rare in direct one-on-one therapy but appear in group, classroom, and token economy contexts where timing of behavior matters — transitions, waiting, scheduled task completion.
Variable Interval (VI)
Reinforcement is available after an unpredictable interval elapses, then delivered for the first response. VI schedules produce low, steady rates of responding and are resistant to satiation.
Use VI when:
- Independent work or on-task behavior (staying seated, engaging in leisure)
- Behaviors that should occur consistently across time rather than in bursts
Choosing the Right Schedule: Acquisition vs. Maintenance
The most common schedule error in ABA practice is thinning too fast. Moving directly from CRF to FR-5 in a single session creates ratio strain — a disruption in responding caused by an abrupt increase in the response requirement. Signs of ratio strain include: prolonged post-reinforcement pauses, increases in off-task behavior, emotional responding, or flat-out extinction of the target behavior.
A practical thinning progression for most skill acquisition targets:
- CRF through first mastery criterion
- FR-2 or FR-3 for two to three sessions
- VR-3 as the next stable schedule
- Gradual VR thinning (VR-4, VR-5) while monitoring rate and quality
The guiding question at each step: is the behavior maintaining its quality and rate as the schedule thins? If responding drops, slows disproportionately, or the learner shows affect changes, you thinned too fast. Step back one level and hold longer before advancing again.
For established, generalized behaviors, VR schedules typically provide the best long-term maintenance. FR schedules work but require attention to post-reinforcement pauses. Interval schedules suit time-based or steady-state goals better than rate-building goals.
Three Thinning Mistakes That Undermine Programs
1. Thinning before acquisition is solid
If the behavior isn’t at criterion on CRF, adding a ratio requirement slows acquisition rather than building toward maintenance. Check your data before thinning. Mastery criteria exist for a reason.
2. Running the same schedule across all behaviors in a session
Different targets sit at different stages of the acquisition-to-maintenance arc. Running a freshly taught mand on VR-5 while also running a maintenance target on VR-5 is a mismatch. Know where each target is and schedule accordingly — even if that means managing three or four different delivery rules in a single session.
3. Treating post-reinforcement pauses as a behavior problem
Post-reinforcement pauses are a normal feature of FR schedules. They are not escape, they are not defiance, and they don’t require a consequence. If pauses are functionally disruptive to the session flow, the solution is transitioning to a variable schedule — not correcting the pause itself.
A Quick Reference Table
| Schedule | Response Pattern | Best Used For |
|---|---|---|
| CRF | Every response reinforced | New skill acquisition, re-establishment |
| FR | High rate, post-reinforcement pause | Thinning from CRF, rate building |
| VR | High, steady, extinction-resistant | Maintenance, language, social behavior |
| FI | Scallop pattern (slow → fast) | Time-based behavior, group programs |
| VI | Low, steady, consistent | On-task behavior, independent work |
How Kipr Helps
Schedule selection and thinning look simple in a textbook. In session, the judgment calls come fast: the learner is showing what looks like ratio strain, but is it actually fatigue from a long session? Is the post-reinforcement pause within a normal range, or is something else driving it? Do you step the schedule back, or hold through it?
These are the decisions that separate experienced practitioners from new ones — and they’re hard to develop through coursework or supervision observations alone. Kipr’s AI-simulated practice scenarios give BCBAs and RBTs a safe place to rehearse schedule implementation, work through thinning decisions, and make the mistakes that build clinical judgment before those calls happen with a real client.
Join the Kipr waitlist for early access.