An ABA shaping plan worksheet should preserve each reinforcement criterion and the evidence behind every change. That history is part of the procedure, not administrative debris. If a note records only “working on the next approximation,” a later reviewer cannot tell what counted, whether the response dimension stayed stable, how the step was selected, or whether implementers reinforced according to the current version.

This copyable worksheet keeps the terminal response, current approximation, response distribution, criterion changes, access, integrity, unwanted effects, and qualified decisions in one audit trail. It does not prescribe a ladder. Shaping a duration, force, distance, latency, topography, or other dimension can involve different measurements and interdisciplinary responsibilities.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

Choose one meaningful response dimension

Begin with the client-selected purpose, then describe the terminal response in observable terms. Choose the response dimension that is actually expected to change. Do not quietly switch from duration to topography, add task steps, change a prompt, or thin reinforcement while still calling every change “shaping.”

Plan identityComplete before useClient-selected purpose and practical outcomeTerminal response, including accepted accessible formsDimension being shaped and unit of measurementOpportunity start and stop boundariesCurrent response range or starting approximationOrdinary supports, AAC, assistive technology and positioningRelevant assessment and interdisciplinary recommendationsConsent, assent/dissent and withdrawal processPlan version, author, qualified approver and effective dateSettings, implementers, exclusions and review date

Kincaid's gradual-change taxonomy separates changes in discriminative stimuli, response requirements, and reinforcement. In that framework, shaping can change a response's formal features or another measured dimension. The distinction helps prevent a task-analysis step, prompt fade, or schedule change from being misfiled as a shaping approximation.

Treat the approximation ladder as a versioned hypothesis

A ladder is a reviewable hypothesis about useful steps. It is not a promise that every person will move linearly through them. Define each criterion so independent observers can classify the same response, and preserve a route to hold, return, revise, stop, or refer.

Versioned stepResponse criterionWhat remains acceptable from the prior stepProgrammed consequenceEvidence to review before changeHold, return or stop evidenceStarting approximationCurrent criterionProposed next approximationLater terminal criterion

Avoid labels such as “close enough” unless the measurable boundary is written next to them. A small study comparing vocal shaping and broad reasonable attempts found different acquisition patterns under its specific definitions, participants, and teaching history. It also reported integrity and agreement challenges. Those results support explicit definitions; they do not make its vocal criteria, mastery rule, or teaching arrangement universal.

Keep attempt-level evidence before the summary

The row-level ledger retains the observed response, prompt or ordinary support, current criterion, eligibility decision, delivered consequence, and context. That makes it possible to distinguish a client response from an implementation mismatch.

Date/timeCriterion versionOpportunity valid?Observed response and measured valuePrompt exposure or ordinary supportCriterion categoryConsequence eligible?Consequence delivered?Assent/dissent, access or context noteYes / No, reasonCurrent / prior / other / no responseYes / No / UnclearYes / No / Unclear

Keep invalid, interrupted, unavailable, device-failure, and not-observed states out of the performance denominator. Do not recode a response that meets an earlier approximation as a generic error; its location in the response distribution can matter when reviewing step size. Prompt exposure also stays visible even when the measured response reaches the current criterion.

Two measures answer two different questions

The fictional example concerns a client-selected goal of pressing and holding an already-selected adaptive switch long enough to activate a chosen music clip. Device fit, placement, calibration, motor access, and positioning remain governed by the applicable interdisciplinary plan. The terminal duration is four seconds. The current version reinforces holds of at least two seconds and less than three seconds.

Across 12 valid attempts, seven meet the current two-second criterion, three meet only the prior one-second step, and two have no switch contact within the defined opportunity.

This filled summary shows the response distribution used for the calculations. The underlying ledger still carries the measured duration and consequence decision for each attempt.

Response category under version 3Valid attemptsConsequence eligible under current criterionConsequence decision matched written ruleCurrent criterion: at least 2 and less than 3 seconds776Prior approximation: at least 1 and less than 2 seconds302No switch contact within the opportunity202Reconciled total12710

  • Current-criterion attainment is 7 / 12 × 100 = 58.3%.
  • Consequence decisions match the written criterion on 10 of 12 attempts, or 10 / 12 × 100 = 83.3%.

The two integrity mismatches are preserved: one eligible response does not receive the programmed consequence, and one response below the current criterion does. The 58.3% value describes the recorded response distribution. The 83.3% value describes implementer decisions. Neither establishes readiness to advance, acquisition, treatment effect, motivation, device suitability, assent, or benefit.

Use the ABA shaping plan worksheet to keep client performance separate from criterion-consequence integrity. Review the attempt rows before interpreting either percentage, especially when the response distribution is clustered near a boundary.

The BACB Sixth Edition BCBA Test Content Outline includes shaping, differential reinforcement, measurement, integrity, and data-based modification as examination content. It does not choose a criterion or threshold. A review of mastery criteria and maintenance reporting supports stating exactly what was measured and followed, not importing one paper's criterion into every case.

Let the whole distribution inform step size

Do not decide from a percentage alone. Look at how responses cluster, whether values are trending or variable, how often prior approximations still occur, and whether measurement remained feasible. Consider implementation fidelity, prompt exposure, reinforcement delivery, device reliability, health, context, and the person's direct response to the activity.

Review questionEvidence and datesWhat the evidence can supportLimits or missing informationOwner/next actionIs the response dimension still the one intended?Are category boundaries observable and applied consistently?How are values distributed within and across sessions?Did prompts, devices, positioning or materials change?Were programmed consequences delivered as written?What assent, dissent, burden, avoidance or unwanted effects occurred?Are interdisciplinary or safety questions unresolved?

A larger step may produce fewer eligible responses; a very small step may fail to create a meaningful change. Neither observation supplies a universal rule. A qualified reviewer needs the actual distribution and the client's experience, not a mechanically increasing ladder.

When a criterion changes, start a new version with its effective date, owner, boundary definition, consequence rule, and reason for the change. Never apply a later threshold backward to earlier attempts. Cross-version summaries should display each version separately before any explicitly justified combined view, so a denominator does not mix unlike criteria.

Communication and access are never shaping contingencies

The ASHA AAC Practice Portal states that AAC users should always have access to their communication tools or devices. Do not make access to communication, food, water, bathroom, breaks, emergency support, mobility, sensory regulation, or medical care contingent on meeting a shaping criterion.

The adaptive switch in the fictional example is part of an access arrangement, not just a mechanism for delivering music. Its suitability may involve occupational therapy, speech-language pathology, assistive-technology, medical, educational, caregiver, and client expertise. A data form cannot settle those roles.

Version every exception and clinical decision

DateException or review triggerEvidence inspectedClient/caregiver inputOptions consideredDecision and rationaleNew version/effective dateQualified ownerContinue / hold / revise / return / stop / refer

A withdrawal response, health change, motor-access concern, sudden performance shift, repeated implementation mismatch, or unstable measurement deserves its own record. It should not be renamed noncompliance or buried in an average. If the criterion changes, issue a new version and leave earlier rows attached to the version under which they occurred.

Covered certificants can consult the live BACB Ethics Codes hub and Ethics Code for Behavior Analysts when reviewing competence, dignity, client participation, risk, documentation, and evaluation limits. Current examination materials are identified on the BACB test-content hub. The CASP ABA Practice Guidelines access page describes Version 3.0 and its licensing conditions; this article neither reproduces licensed detail nor claims compliance.

An assent-focused practice article can inform questions about observable assent and withdrawal, but it does not create a universal code or replace applicable consent law and organizational requirements. Use the worksheet alongside the current assessment, treatment plan, consent and assent process, safety plan, authorization, payer rules, law, and interdisciplinary recommendations.

Related resources

Sources