Use an ABA error correction procedure worksheet to make a teaching pathway inspectable. Record what happened before an error, what counted as the error, what the clinician planned next, how the learner responded, and when the sequence stopped. The record should preserve meaningful differences among events rather than turn every incorrect response into the same story or every correction into a universal recipe.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

The question this record should answer

Error correction is not a single procedure. A written plan may include a model, a prompt, an active learner response, a repeated opportunity, feedback, reinforcement, or another arrangement. The right combination depends on the skill, the learner, the teaching context, assessment evidence, communication access, and the effects observed during implementation.

The BACB Ethics Code places intervention selection within scientific evidence, assessment results, client and stakeholder needs, risk reduction, written description, correct data collection, and continual evaluation. The current BCBA Test Content Outline likewise distinguishes prompting, prompt fading, reinforcement, discrimination procedures, unwanted effects, procedural integrity, and data-based modification. Neither source supplies a required error-correction sequence.

Use this working artifact to answer a narrower question: Can another qualified person understand the planned teaching episode, reproduce the approved pathway, and see how the learner responded? It does not determine whether the program is medically necessary, whether the target is socially important, or whether a particular correction is clinically appropriate.

The worksheet is descriptive, not self-authorizing. A blank field should stay blank until the relevant evidence or observation exists.

The teaching episode before the error

A correction cannot be interpreted apart from the opportunity that preceded it. Record the target, materials, instruction or relevant cue, expected response, response window, prompting arrangement, and what makes an opportunity eligible. If the learner could not access the materials, did not receive the instruction, lacked the agreed communication support, or withdrew participation, do not relabel the event as an error.

That distinction protects the denominator. It also preserves the learner's response to the teaching context.

Program fieldCase-specific entryClient purpose and current goalObservable target responseMaterials and settingInstruction or antecedentIndependent response windowCorrect, incorrect, and no-response definitionsPrompting arrangement before correctionEligible-opportunity ruleCommunication and access supportsCurrent consent, assent, and plan reference

An operational definition belongs here, not in a staff member's memory. It should be precise enough to distinguish an incorrect response from a response that is different but acceptable, a response using AAC, a delayed response within the allowed window, a request for help, or an unavailable opportunity.

The approved correction pathway

Write the sequence as observable actions and decision points. Avoid labels such as “standard correction” or “repeat until correct,” because they conceal timing, response requirements, limits, and consequences.

Each branch needs its own exit. The endpoint should be observable to someone reading the current plan.

StepTriggerStaff actionLearner opportunityProgrammed consequenceExit or stop condition1. Detect and verify2. Provide the approved cue, model, or prompt3. Offer an active response opportunity, if applicable4. Re-present or test later, if applicable5. Close the episode

The form deliberately leaves the sequence blank. Research has compared multiple error-correction and errorless arrangements, but effects can differ across learners, targets, and procedures. A randomized comparison of error correction and most-to-least prompting found both effective for the study's tact targets, with differences on some outcomes. That finding does not establish one default for every learner or skill. An assessment-based instruction tutorial similarly emphasizes evaluating alternatives rather than assuming that a generally supported method is the most efficient or acceptable option for an individual.

First and corrected responses answer different questions

An independent correct response on the first opportunity and a correct response after a model or prompt answer different questions. Combining them can make acquisition look stronger than the first-response record supports. It can also hide whether correction episodes are becoming shorter, more frequent, or more dependent on a particular prompt.

Those measures belong in different columns. Their trends may also call for different clinical questions.

EventInitial responseCorrection pathwayResponse after correctionAccess, dissent, or stop eventNotesIndependent correct / incorrect / no response / invalidNot entered / completed / modified / stoppedIndependent / prompted / incorrect / no response / not reached

A corrected response can be useful evidence about the immediate sequence. It is not retroactive independent accuracy. If a later unprompted probe is important, give that probe its own row, antecedent, timing, and response classification.

Learner access and response within the procedure

Communication access affects what a teaching opportunity means. ASHA's Communication Access guidance recommends asking about communication preferences, communicating directly, supporting multiple methods, allowing processing time, and checking understanding. In this worksheet, those practices translate into explicit fields for the learner's established mode, processing interval, help or break response, accessibility supports, and the staff response to withdrawal behavior.

Do not classify dissent, a request to stop, distress, pain, equipment failure, an inaccessible stimulus, or a missing prerequisite as an incorrect target response. Follow the current individualized plan and applicable safety, consent, and escalation process. A documented stop may show that the boundary worked as intended; it is not automatically missing treatment integrity.

A pause is data, not permission to continue. Its meaning still requires case-specific review.

Boundary or access eventObservable descriptionImmediate responseWas the teaching event still eligible?Follow-up owner and dateWithdrawal, pause, or stop responseYes / no / unresolvedPossible distress or unwanted effectCommunication or AAC access problemMedical, sensory, or environmental concernProcedure exception

Implementation review without a single-score shortcut

Treatment integrity should identify which components were and were not delivered. A total percentage alone can hide a missed stop condition, a changed prompt, an absent reinforcer, or a timing error that materially altered the episode.

Component-level notes keep that difference visible.

Integrity componentPlanned evidenceObserved as writtenDifference or not-applicable rationaleActionEligible opportunity verifiedYes / no / unknownInitial response classified before correctionApproved correction step deliveredPrompt and timing matched the current planProgrammed consequence deliveredAccess and stop conditions honoredCorrect data stage recorded

Peer-reviewed component analyses of skill-acquisition consequences show why these distinctions matter. Studies have examined active-response correction, response repetition, re-presentation, and reinforcement after correction as separable elements. Findings from a small study or one target class cannot settle which components should be used elsewhere. They can support a habit of naming the components and evaluating them rather than treating “error correction” as a complete description.

A fictional completed example

The following illustration uses invented data for a fictional receptive-label program. It is not a recommended sequence or criterion.

The current plan defines 12 eligible first opportunities. Eight end with an independent correct response and four with an initial error. For the four errors, three approved correction pathways are completed. On the fourth, the learner uses the established stop response before correction begins; staff end the teaching episode and record follow-up.

Event groupCountWhat the count meansEligible first opportunities12The instruction, materials, access, and response window were availableIndependent correct first responses8Correct before the correction pathwayInitial errors4Met the current operational definition on the first opportunityCompleted correction pathways3Approved pathway reached its documented endpointStopped pathway with reason recorded1Withdrawal response was honored; not scored as a completed correction

Initial independent accuracy is:

8 / 12 × 100 = 66.7%

Error-pathway documentation is:

4 / 4 × 100 = 100.0%

The second calculation counts three completed pathways plus one stopped pathway with its reason and response documented. It does not call the stopped event a successful correction. The two percentages should not be added or averaged. They do not prove acquisition, maintenance, generalization, preference, comfort, procedural effectiveness, or readiness to continue unchanged.

The denominator remains the four initial errors. No outcome is inferred from documentation completeness alone.

Questions raised by the pattern

An ABA error correction procedure worksheet becomes useful when it leads to a qualified review, not when it generates a verdict. Consider the pattern across time and conditions. Are first-response errors clustered around a stimulus feature, missing prerequisite, prompt transition, implementer, setting, pace, or access barrier? Are correction episodes followed by independent responding on later opportunities? Are unwanted effects, avoidance, withdrawal, or help requests changing? Is the written procedure actually the one being delivered?

Record uncertainty directly. “Not observed,” “not applicable,” “invalid,” and “pending review” should remain distinct. If the observed pathway departs from the current plan, pause interpretation until the difference is understood.

Review questionEvidence reviewedCurrent interpretationDecision ownerNext reviewContinue unchanged, modify, compare an alternative, pause, or stop?Is additional assessment or consultation needed?What would change the decision?

Related resources

Sources

The CASP access page describes the purpose and licensing pathway for its current practice guidelines. This original worksheet does not reproduce licensed guideline text or claim to be a CASP form.