Variable-ratio schedule

Response count changes across ratios

Each completed ratio produces the scheduled consequence and begins the next requirement. If the sequence is 2, 3, 5, 4, and 6, completing 20 eligible responses produces five outcomes. The average ratio is 20 divided by 5, or four.

Specify what qualifies. A completed package, accurate data entry, or independent communication can be counted. Partial attempts, prompted responses, and corrections need explicit rules.

VR differs from fixed ratio

A fixed-ratio schedule uses the same requirement every time. FR 4 requires four responses for each outcome. VR 4 changes individual requirements while maintaining an average of four across the planned distribution.

VR also differs from variable interval. VI changes elapsed time before the next response becomes eligible. VR changes response count.

Baum’s schedule analysis reviews ratio and interval definitions and performance differences. It is a conceptual account rather than a clinical recommendation.

The average never describes the whole schedule

A 3, 4, 5 sequence and a 1, 4, 7 sequence both average four. The second has a wider range and may feel less predictable. Record all values, their frequency, and the realized order.

When software samples ratios, preserve the generated sequence or seed when practical. Compare the programmed and delivered distributions. Session endings can leave incomplete ratios and shift the realized average.

A fictional VR record

Eren chooses a short quality-check summary after varying package counts. Six requirements are 2, 3, 5, 4, 6, and 4. They total 24 responses, so the mean ratio is 4. Eren completes all six ratios and receives six summaries.

Implementation is 6 of 6 completed ratios. The arithmetic confirms the planned sequence; it does not show that VR improved accuracy, speed, satisfaction, or maintenance. Package difficulty and response time also varied.

Response patterns are outcomes

VR schedules often support persistent responding in laboratory arrangements because the next requirement is uncertain. Human instructions, task effort, competing activities, consequence quality, and ratio range can produce different patterns.

Staddon and Cerutti review operant schedules broadly. Lattal and colleagues’ ratio-pausing review shows how upcoming requirements and correlated stimuli can affect pauses. Neither source establishes a desirable clinical rate.

Measure integrity and burden separately

Report scheduled outcomes delivered divided by completed ratios. Also track eligible responses, incomplete ratios, errors, prompts, pauses, time per ratio, consequence delay, preference, and fatigue.

Do not hide unfinished work. If a session ends two responses into a five-response requirement, record the partial ratio and whether the next session carries it forward.

Clinical use requires clear boundaries

Explain the changing requirement in an accessible way. Some people prefer a visible progress counter; others may find variable requirements frustrating. Assess fit and offer alternatives.

Food, water, bathroom access, AAC, mobility, prescribed care, rest, and emergency help remain available according to need. A VR schedule cannot make basic access uncertain.

The BACB BCBA Test Content Outline, 6th edition includes reinforcement schedules and thinning in examination content. It does not prescribe ratios or a person-specific plan.

Version each change

Changing the list, mean, range, response definition, or consequence creates a new schedule version. Mark its effective date and graph it separately. When thinning, use predeclared decision rules and monitor burden alongside performance.

Read the schedule from an event record

A useful VR record shows the sequence of requirements rather than only the session average. Suppose a practice programs ratios of 2, 5, 3, 7, and 3 responses. The mean is four, but the person experiences five different requirements. Record each completed ratio, the response count, elapsed time, consequence delivered, prompts, pauses, and any early stop. That sequence reveals whether a long requirement preceded a pause or whether delivery drifted away from the programmed values.

Before interpreting a response rate, ask three practical questions:

• Did every eligible response use the same operational definition?
• Did staff deliver the consequence immediately after the programmed count?
• Did the person have comparable access, health, motivation, materials, and communication support across observations?

Audit schedule integrity as completed ratios matching the programmed requirement divided by completed ratios reviewed. Keep that percentage separate from the person’s response rate and from whether the arrangement was acceptable to the person. A high response rate can coexist with delivery errors or excessive effort. Accurate implementation also leaves the value of the selected goal and consequence as a separate question.

When performance changes, examine the event record before raising the average ratio. A schedule adjustment should have a stated purpose, qualified clinical review, and a plan to check effort, choice, generalization, and side effects.