An ABA time delay procedure worksheet should show the programmed interval between a cue and a prompt, the delay step actually used, and whether the learner responded before or after the prompt. A session-level accuracy score cannot reveal those distinctions.

Use this ABA time delay procedure worksheet to document a case-specific constant or progressive delay arrangement, keep invalid opportunities visible, and review independent-before-prompt and prompted-correct responding separately. The fictional example checks the arithmetic without prescribing a delay schedule.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

Clinical boundary: Time delay belongs inside an authorized, individualized process led by qualified professionals. Maintain ordinary speech, AAC, gesture, sensory, motor, visual, medical, safety, relationship, and processing-time supports. Respond accessibly to help, break, stop, refusal, pain, and alternative communication. This worksheet cannot establish consent, assent, medical necessity, effectiveness, mastery, payer coverage, or legal compliance. See current BACB ethics resources for professional context.

A delay is a programmed relation, not waiting in general

Time delay changes when a controlling prompt is scheduled after a defined cue. Constant time delay typically begins with an identified zero-delay teaching phase and then uses a fixed nonzero delay. Progressive time delay increases the interval according to a case-specific rule. Simply pausing, waiting for attention, allowing ordinary processing time, or delivering a late prompt is not enough to identify either procedure.

The record therefore needs a cue onset, scheduled delay, actual prompt time, initial response time, and outcome code. If the learner responds correctly before the prompt, that may count as independent under the stated definition. If the prompt occurs first, the later correct response remains prompted. If timing cannot be reconstructed reliably, mark the opportunity invalid or timing-unknown under the plan rather than guessing.

Sequence matters here. The timestamp gives the code its meaning.

Research has compared constant time delay with simultaneous prompting, constant with progressive delay, and progressive delay in specific early-learning applications. These small studies differ in participants, targets, prompting arrangements, and outcomes. They justify naming and measuring the actual procedure, not choosing a universal delay.

Purpose and procedure card

Define the meaningful skill first. Then write enough procedure detail for a reviewer to distinguish a planned delay from ordinary latency, timer drift, or a different prompting method.

FieldCase-specific entryMeaningful skill purpose and authorized plan referenceTarget, cue, materials, setting, and natural contextObservable initial response and latency ruleAccessible response forms that count as equivalentControlling-prompt assessment and selected promptConstant or progressive delay rationaleZero-delay history and transition evidenceIndependent-before-prompt, prompted-correct, error, no-response, and other codesConsequence and correction rulesValid-opportunity and timing-unknown rulesHelp, break, refusal, alternative, and stop responsesQualified owner, version, effective date, and review date

Avoid turning response speed into a value judgment. A longer latency may reflect motor planning, communication access, unfamiliar materials, pain, environmental distraction, or the person's deliberate pace. The initial response interval should be clinically justified and accessible, not selected to make independent responding look larger.

Speed is not the goal by default. Meaningful access comes first.

Versioned constant-or-progressive delay map

Write one row for every procedure version or step. A constant-delay record needs the fixed interval actually used. A progressive-delay record needs the sequence and advancement or return rule. Both need a zero-delay description when that phase is part of the plan.

Version/stepDelay typeScheduled delayCue and timer startPrompt topographyResponse-window boundaryAdvancement/return ruleConsequence/correctionAccess supportsEffective datesOwnerzero-delay teaching0 secconstant delayprogressive delay

A foundational review of constant time delay describes the zero-delay-to-fixed-delay logic found in that literature. It is useful historical context, not a contemporary protocol for a specific learner. A delay established in an article is not automatically accessible, ethical, safe, or effective in a different service.

Do not change delay steps informally while keeping the old label. Record who authorized the change, what data were reviewed, when it took effect, and what happens after errors, no responses, access interruptions, distress, or repeated timing mismatches.

One version should describe one rule. Retire ambiguity, not history.

Opportunity ledger: preserve actual timing

One row represents one planned opportunity. Use a clock, timer, video permitted under the privacy plan, or another defined measurement method. Precision should match what the implementation system can support; false decimals do not improve a rough observation.

Date/timeVersion/stepTarget/cuePlanned delayActual prompt timeValid? reasonInitial response timeBefore/after promptOutcomeCorrection/supportHelp/refusal/breakAccess/contextImplementeryes / nobefore / after / none / unknownindependent correct / prompted correct / incorrect / no response / otheryes / noyes / no

If the prompt is early or late beyond the defined tolerance, keep the observed learner response but mark the implementation mismatch. If the cue was unclear, the timer did not start, materials were wrong, or the required communication access was missing, record invalidity. Never convert a provider-side problem into an incorrect learner response.

The row can preserve both facts. It need not choose one.

Delay-step summary

Overall totals can mask the central time-delay question. Preserve outcomes at every step before calculating a cross-step descriptive total.

Version/stepPlannedInvalidValidIndependent before promptPrompted correctIncorrectNo response/otherMedian/range latencyTiming mismatchNotes0 seconds

At zero seconds, independent-before-prompt may be structurally unavailable because the prompt is scheduled immediately. Do not interpret a zero in that column as failure. Its purpose may be error-minimizing teaching or establishing the cue-prompt relation under the individualized plan. A later nonzero step asks a different question.

An overall “correct” total may combine independent-before-prompt and prompted-correct responses for a limited implementation question. It must never replace the separate columns. Also preserve errors and no responses by step; longer delays can change both opportunity for independence and opportunity for error.

Step totals show where that change occurred. Overall totals cannot.

Access, learner experience, and burden

The timer should not outrank the person. Communication access, sensory or motor support, ordinary processing time, and a safe way to ask for help remain available. If those supports change the condition, document the change; do not remove them solely to preserve a nominal delay.

Review dateVersion/stepCommunication/access availableLearner choice or descriptionHelp/break/stop honored?Fatigue, pain, distress, or sensory effectTiming or prompt burdenRequested changeFollow-up owner

The ASHA AAC Practice Portal describes AAC as individualized, multimodal, and flexible. A response on a device may require navigation or partner support that should be reflected in the response definition and interval. Correctness and speed cannot establish assent, preference, comfort, or benefit.

Record the person's own communication. Do not replace it with inference.

Separate denominators before interpreting change

Review questionNumeratorDenominatorResultDoes not establishValid opportunity coveragevalid opportunitiesplanned opportunitieslearner accuracyIndependent-before-prompt rateindependent correct before promptvalid opportunitiescausation or generalizationPrompted-correct ratecorrect after promptvalid opportunitiesindependenceCorrect by either routeindependent plus prompted correctvalid opportunitiesindependent acquisitionPlanned independent yieldindependent correct before promptall planned opportunitieswhy opportunities were lostStep-specific error rateerrors at stepvalid opportunities at stepsafety or treatment failureTiming integrityprompts delivered within defined timingobservable prompt opportunitieslearner benefit or assentFull implementation fidelitycorrectly implemented componentsobservable component opportunitiesmastery, preference, or coverage

Carry raw counts next to percentages. A change from zero to six independent responses looks different if it occurred across ten opportunities than across one hundred. It also means little without the step, sequence, target set, prompt, access conditions, and prior exposure.

Context makes the percentage interpretable. Keep the context visible.

Fictional community garden-label example

This fictional client helped select a community-garden activity involving labels for shared tools. The team planned 30 opportunities across sequential 0-, 2-, and 4-second steps. The example demonstrates coding and arithmetic only. It does not endorse the target, sequence, or delays.

Three opportunities were invalid because the timer failed once, the wrong cue was presented once, and the learner's communication display was unavailable once. Twenty-seven valid opportunities remained, so valid coverage was 27 / 30 × 100 = 90.0%.

At zero seconds, eight opportunities were valid. None could be scored independent before the immediate prompt; seven were prompted correct and one had another outcome. At two seconds, 10 were valid: six independent before the prompt, three prompted correct, and one other. At four seconds, nine were valid: seven independent before the prompt, one prompted correct, and one other.

Delay stepValidIndependent before promptPrompted correctOtherBounded reading0 seconds8071Immediate prompt makes the independent-before-prompt route unavailable2 seconds10631Describes this step after prior exposure4 seconds9711Describes this step after earlier stepsTotal2713113Step-level data remain visible

Across valid opportunities, independent-before-prompt responding was 13 / 27 × 100 = 48.1%. Prompted-correct responding was 11 / 27 × 100 = 40.7%. Correct by either recorded route was 24 / 27 × 100 = 88.9%. Planned independent yield, which keeps invalid opportunities visible, was 13 / 30 × 100 = 43.3%.

The increasing count across delay steps does not prove that increasing delay caused improvement. The steps occurred in sequence, the samples are small, prior teaching accumulated, and context may have changed. The 88.9% combined correct figure is not an independence score.

It is a bounded implementation summary. Read the step rows first.

Fidelity and timing record

ObservationCorrect version availableCue/timer accurateDelay step accuratePrompt timing in toleranceOutcome coded by timingConsequence/correction accurateAAC/access preservedHelp/refusal honoredCorrect componentsOpportunitiesNotes

In the fictional sample, 15 observations across eight components created 120 component opportunities. There were 111 correct components and nine mismatches, which sum to 120. Fidelity was 111 / 120 × 100 = 92.5%. Review the nine mismatches by component and step. Timing drift at four seconds has a different implication than an inaccessible response mode at zero seconds.

A brief assessment of prompting procedures and research comparing instructional prompting procedures illustrate why learner-specific efficiency and error patterns matter. Those studies do not replace a full assessment, direct learner feedback, maintenance checks, or generalization evidence.

Brief comparisons answer brief questions. Follow-up still matters.

Maintenance, generalization, and decision record

FieldReview entryVersions, steps, targets, settings, partners, and dates reviewedMeaningful purpose and continued authorizationPlanned, invalid, and valid opportunitiesIndependent, prompted, error, no-response, and latency patterns by stepTiming and full-procedure integrity limitsMaintenance interval and intervening exposureGeneralization examples, people, settings, or response formsAccess, assent/dissent, choice, burden, and unwanted effectsRetain / revise / reassess / fade / return / pause / stop questionQualified owner, rationale, version, and next review date

The Ethics Code for Behavior Analysts addresses rights, involvement, competence, confidentiality, documentation, and responsible data use. The current BACB test-content outlines and Sixth Edition BCBA outline cover prompting, prompt fading, skill acquisition, integrity, maintenance, generalization, and data-based decision-making. The CASP ABA Practice Guidelines access page supplies practice context. None selects a delay or certifies a case decision.

Reopen the procedure when timing cannot be implemented reliably or errors and no responses concentrate at a step. Reopen it when access differs by step, learner communication changes, burden increases, the purpose no longer holds, or follow-up does not support the intended outcome. The qualified team may revise the response window, assess another prompt, change the sequence, return to an earlier step, pause, or stop. The worksheet makes the evidence reviewable; it does not automate that judgment.

Document the rationale. Set a real review date.

Related resources

Sources

The sources differ in population, target, prompt, delay, teaching history, consequence, and design. They support explicit timing, step-level outcomes, access, direct measurement, fidelity checks, and careful interpretation. They do not define a universal zero-delay duration, delay type, increment, maximum delay, response interval, trial count, reinforcement, correction, mastery rule, or expected effect.