An ABA skill acquisition data sheet should make each recorded response understandable without forcing the next reviewer to guess which target, prompt, opportunity, or program version was in effect. It should also keep session arithmetic separate from the larger clinical judgment about whether instruction should continue, change, generalize, move to maintenance, or pause.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

This copyable worksheet pairs a trial-level record with a mastery-review panel. It does not choose a teaching procedure or mastery rule. A qualified clinician must define those parts in the current plan, train implementers, and decide how the resulting evidence will be used.

Start with a versioned measurement contract

Record the rules before recording responses. If the team changes a response definition, prompt code, opportunity rule, or teaching procedure, open a new version rather than silently combining unlike data.

Program identityEntryClient or authorized case code________________.Goal and target name________________.Program version and effective date________________.Responsible clinician________________.Setting, materials, people, and relevant context________________.Observable target response________________.What counts as an eligible opportunity________________.Independent-response definition________________.Prompt codes and hierarchy, if applicable________________.Error, no-response, blocked-trial, and exclusion codes________________.Teaching procedure authorized in the current plan________________.Current mastery or transition rule and source________________.Generalization and maintenance plan________________.Client communication, choice, assent, and dissent plan________________.Safety and immediate stop conditions________________.Data storage, access, correction, and retention rule________________.

Read the entries back before use. Resolve conflicting definitions. Train each assigned implementer. Record the effective version in the same system.

The current BACB Ethics Code hub points readers to the controlling code. The Ethics Code for Behavior Analysts, updated August 2024 when checked, addresses competence, confidentiality, records, assessment, behavior-change programs, consent, client welfare, and data accuracy. This worksheet does not determine whether a particular program satisfies those obligations.

A blank trial record

Use one row for one defined eligible opportunity. Record the initial response before a correction or retry. If a planned opportunity cannot occur, use an exclusion code with a reason; do not turn it into an error merely to preserve a denominator.

OpportunityTime or routineTarget or exemplarInitial responsePrompt levelError or no responseCorrection or retry outcomeClient choice, assent, or dissentInclude in denominator?Context note1.2.3.4.5.6.7.8.

Review the sheet before leaving the setting. Mark late entries as late. Preserve unclear events as unclear. Ask rather than guess.

Add rows as needed, but do not manufacture extra trials to make a percentage look stable. When teaching and probes use different consequences or prompting rules, identify them as different conditions. A successful corrected response may be useful teaching information, but it is not the same event as an independent initial response.

Session summary and transparent calculations

Keep counts and denominators visible. A percentage without the opportunity rule can conceal excluded trials, unequal exposure, or changes in scoring.

Session measureCount or calculationPlanned opportunities________________.Eligible opportunities________________.Excluded or blocked opportunities, with reasons________________.Independent initial responses________________.Prompted initial responses________________.Errors________________.No responses________________.Corrections or retries completed________________.Independent percentageIndependent initial responses / eligible opportunities × 100.Teaching exposureTrials, minutes, exemplars, people, settings, or other plan-defined unit.Treatment-integrity evidence availableYes / no / partial; source and date.Data-quality concernMissing / late / changed definition / unclear code / observer concern / none noted.

Check the arithmetic twice. Keep source rows available. Note any missing trials. Retain the reason for each exclusion.

Do not collapse counts that answer different questions. Independent percentage describes performance under its definition. Prompt distribution describes support used. Treatment integrity describes implementation. Actual exposure describes how much instruction or practice occurred. None alone proves learning, durability, generalization, social validity, or treatment effectiveness.

A mastery review should slow the decision down

Research does not support treating one common percentage as a universal mastery rule. A descriptive review of applied skill-acquisition studies found substantial variation and incomplete reporting in mastery and maintenance practices (McDougale et al.). Experimental work has examined how criterion level can affect maintenance, while also calling for more research and individualized interpretation (Fuller and Fienup). A more recent replication compared criteria applied to individual operants and sets, with participant-level variation in maintenance outcomes (Cordeiro and colleagues). These studies inform questions; they do not supply a default criterion for every client, skill, or risk level.

Review questionEvidence and answerIs the authorized criterion stated with level, unit, observations, and conditions?Were enough eligible opportunities actually observed under the defined condition?Are the current and prior sessions comparable?Did prompts, materials, people, setting, task difficulty, or motivation change?Are treatment-integrity and observer-quality data adequate for this decision?What does the client communicate about acceptability, effort, or stopping?Is performance represented across relevant exemplars, people, settings, or routines?Is maintenance evidence required and available?Are safety, health, access, AAC, sensory, or contextual factors unresolved?What evidence is missing or uncertain?Who has authority to continue, revise, probe, generalize, maintain, or pause?Decision, rationale, effective date, owner, and next review trigger

Name the decision owner. Record disagreement. Flag conflicting evidence. Set a review trigger. Do not turn silence into agreement.

The CASP ABA Practice Guidelines access page identifies Version 3.0, released in 2024, and its licensing conditions. This article links to the access page without reproducing licensed content. Use the current licensed guidance, treatment plan, payer requirements, law, and organizational policy directly when they apply.

Fictional completed example

The following data are invented for training. Case R is not a real client. The target, codes, criterion, counts, and decision are not recommendations.

Program identityFictional entryTargetCase R independently selects the pictured item named in a two-item arrayVersionProgram 3, effective September 1, 2026Eligible opportunityBoth pictures visible, attending response observed, instruction delivered once, no documented interruptionIndependent responseCorrect selection within the plan-defined interval before a promptCodesI independent; P prompted; E error; X excluded; C correction completedAuthorized review ruleUse the criterion already documented in Program 3; this worksheet does not restate it as a recommendationClient safeguardHonor the documented break signal; do not score a break as an error

Across the fictional session, 26 opportunities were planned. Two were excluded: one because a picture was missing and one because the client used the documented break signal before the opportunity began. The eligible denominator is therefore 24, not 26. Initial responses were 18 independent, 4 prompted, and 2 errors. The descriptive independent result is 18 divided by 24 = 75.0%. Eighteen plus four plus two equals all 24 eligible opportunities.

Evidence streamFictional resultIndependent initial responses18 / 24 = 75.0%Prompted initial responses4 / 24 = 16.7%Errors2 / 24 = 8.3%Excluded opportunities2, kept outside the denominator with reasonsCorrectionsBoth error trials ended with a corrected response; those corrections were not recoded as independentTreatment integritySeparate observation sampled 7 of 8 eligible procedure steps; one material-position step differedClient experienceBreak signal honored; no conclusion about acceptability outside this sessionGeneralizationNot assessed in another person, setting, or exemplar setMaintenanceNot assessed

The fictional reviewer does not mark the target mastered. The record shows only one session, an implementation difference, and no generalization or maintenance evidence. The limited decision is to retain Program 3, correct the material-position support, collect the next planned comparable sample, and review sooner if the client withdraws assent, a safety concern appears, or the definition cannot be applied reliably.

Version and correction log

Never overwrite a definition while leaving old data labeled as though nothing changed.

Version or correction dateWhat changedReasonAffected data rangeAuthorized byRecalculation neededPeople notifiedYes / no / unknownYes / no / unknown

A correction should preserve the original entry, the corrected value, the reason, author, date, and effect on summaries under the organization's record rules. Do not backdate a program version or alter a denominator merely to make a criterion appear met.

Privacy and record boundaries

The HHS Privacy Rule overview explains the federal framework for covered entities and protected health information, including minimum-necessary concepts where applicable. The HHS Security Rule risk-analysis guidance describes risk analysis as foundational to security compliance. They do not approve a particular spreadsheet, device, message, storage location, disclosure, or retention period.

Store the worksheet only in an authorized record system. Confirm access, correction, signature, retention, export, backup, and disposal rules. A working data sheet does not replace a clinical note, treatment plan, consent, incident report, supervision record, authorization submission, claim, or required electronic health record entry.

Stop conditions

Pause data collection when the target, program version, opportunity rule, response code, responsible clinician, client authorization, safety procedure, or record destination is unclear. Stop the teaching sequence and use the applicable clinical or emergency route for an immediate safety concern, acute health change, unexpected adverse effect, repeated distress, assent withdrawal, inaccessible communication, or implementation outside the person's competence.

Do not infer mastery from one percentage, count corrected responses as independent, score blocked opportunities as errors, combine probe and teaching data without labels, or treat staff attendance as implementation fidelity. Preserve uncertainty and obtain qualified clinical review before changing care.

Related resources

Sources