An ABA permanent product data sheet records a lasting result only after the team has defined the product, the opportunity that should produce it, the review window, the scoring rubric, and the evidence connecting it to the person and response. Unavailable, unattributed, and unscorable products should remain visible. Availability, valid scoring coverage, and performance each need a separate denominator.

Clinicians & ABA Professionals / Data, Outcomes and Clinical Decision-Making.

This working template is for a product that remains available after the response, such as an assembled item, completed form, saved file, or dated system entry. The form cannot decide whether that product is a valid stand-in for direct observation. A qualified clinician must make that choice from the measurement question, the current plan, the setting, the person's communication and preferences, and the likely sources of error.

Decide what the product can actually show

The current RBT Test Content Outline, Third Edition lists permanent-product recording separately from continuous and discontinuous measurement. The current BCBA Test Content Outline, Sixth Edition distinguishes direct, indirect, and product measures and asks practitioners to consider validity, reliability, the critical dimension of behavior, and environmental constraints. These outlines describe certification content. They do not validate a product, rubric, or clinical use.

Before adopting product measurement, ask whether the result has a dependable connection to the response of interest. A completed document may show that the document exists. It may not reveal who created it, how much help was used, whether an automated process filled fields, how long the work took, what the person communicated, or what happened during the response.

Use direct observation or another measure when the process matters. Product review may still complement that evidence, provided its limits stay explicit.

Write the product contract before collection

The table works best when it is complete before the first product is reviewed. A change to the definition, rubric, attribution rule, storage location, or review window calls for a new version.

Product contractEntryClient or authorized case codeGoal and measurement questionObservable response of interestDefined product and acceptable formsEligible opportunity and due pointReview window before the product may changeProduct location or authorized record systemAttribution evidence requiredPermitted help, prompts, collaboration, or automationUnavailable, missing, altered, duplicate, and unscorable codesRubric version and effective dateScoring unit and denominator ruleIndependent-rescore plan, if applicableClient communication, choice, assent, and dissent planAccessibility and ordinary support planSafety and immediate stop conditionsAccess, correction, retention, export, and disposal ruleResponsible clinician and next review trigger

Read the entries back to the people who will create, collect, and score products. Resolve examples that fit more than one code. A blank cell is not an instruction to guess.

Blank product register

Each expected product belongs on its own row, including a product that is missing or cannot be scored. Keeping the expected set visible prevents an impressive percentage from hiding poor coverage.

Product IDOpportunity date and due pointProduct found?Location and review timeAttribution evidenceHelp or automationRubric versionScorable?Score or item resultExclusion or concernReviewer1Yes / noYes / no2Yes / noYes / no3Yes / noYes / no4Yes / noYes / no5Yes / noYes / no6Yes / noYes / no7Yes / noYes / no8Yes / noYes / no

Add rows without deleting failed, blocked, or unresolved opportunities. If the product changes after the review window, preserve which version the reviewer scored.

Build a versioned scoring rubric

Define the smallest useful scoring units. A global pass can hide which component was missing, while an excessively detailed rubric can create burden without improving the decision.

Rubric itemObservable pass rulePartial or not-applicable ruleExamples and nonexamplesSafety or access caveatItem 1Item 2Item 3Item 4

Calibrate reviewers on the actual version before use. Record whether a rescore was blind to the original score. Agreement may support scoring consistency for the sampled products, but it is not proof that either score is accurate or that the product belongs to the named person.

Keep the denominators separate

Four different questions sit behind this review, and each needs its own percentage and denominator.

QuantityCalculationWhat it describesProduct availabilityProducts found / expected products × 100Whether expected artifacts were available for reviewAttributable coverageProducts with sufficient attribution / expected products × 100How much of the expected set can reasonably be linked under the stated ruleValid scoring coverageProducts scorable under the current rubric / expected products × 100How much of the expected set supports the planned scoring calculationPerformance among scorable productsProducts meeting the stated criteria / scorable products × 100The scored result for the valid subset, not the whole expected setRescore agreementItem agreements / items independently rescored × 100Descriptive agreement within the sampled rescore

Report counts beside percentages. Explain every excluded product. If the expected count itself is uncertain, stop and resolve the sampling frame rather than presenting a false denominator.

Fictional completed example

The following values are invented for arithmetic review. Case P is not a real client, and the product, rubric, and percentages are not recommendations.

Case P chose a vocational goal involving labeled supply kits. The fictional product is a sealed kit left at an assigned workstation after an eligible routine. Rubric version 2 requires the assigned label, the complete three-item set, and an intact seal. The team expected 12 products during the review period.

Product dispositionCountTreatment in calculationsExpected products12Availability, attribution, and valid scoring denominatorProducts found in the review window10Two expected products remain unavailableProducts with sufficient attribution9One found product has unresolved shared-workstation authorshipProducts scorable under rubric version 28One attributable product was altered before reviewScorable products meeting all three criteria6Performance numeratorScorable products not meeting all criteria2Preserve the failed rubric items

The calculations answer different questions:

  • Availability is 10 divided by 12, or 83.3%.
  • Attributable coverage is 9 divided by 12, or 75.0%.
  • Valid scoring coverage is 8 divided by 12, or 66.7%.
  • Performance among scorable products is 6 divided by 8, or 75.0%.

Reporting only 75.0% performance would conceal that four expected products could not support that score. The record does not treat unavailable products as failures, the unattributed product as Case P's work, or the altered product as if it remained in its original state.

An independent reviewer then rescores four item decisions sampled from the eight scorable products. The two reviewers agree on three items and differ on one, giving 3 divided by 4, or 75.0% item agreement. That small sample describes those four comparisons only. It does not establish scoring accuracy, authorship, independence, or reliability across the full period.

Use permanent-product research within its limits

An early review of academic permanent-product data and reliability procedures documented long-standing measurement and reliability questions in applied research. A later review of school-based treatment integrity describes permanent products as one way implementation has been measured. A recent practice survey of procedural-fidelity data collection reports how respondents used direct, indirect, and product-based approaches in that survey context. These sources support careful method description. They do not establish that product review is valid for every response, person, setting, or clinical decision.

The CASP ABA Practice Guidelines access page identifies Version 3.0 and its licensing conditions. This article does not reproduce the licensed guidelines or claim that the worksheet meets them. Consult the current licensed document and applicable requirements directly.

Preserve communication, access, and privacy

The current ASHA AAC practice guidance provides context for multimodal communication and access. A product created with AAC, adaptive equipment, an interpreter, a visual support, or ordinary assistance should not be downgraded merely because support was used. Record the support and score under the plan-defined rule.

The BACB Ethics Codes hub points readers to current controlling materials. The Ethics Code for Behavior Analysts addresses competence, confidentiality, records, assessment, behavior-change programs, consent, client welfare, and data accuracy. Application depends on role and circumstances, and this worksheet does not provide ethics or legal advice.

Products may contain health, school, employment, location, image, communication, or third-party information. The current HHS Security Rule risk-analysis guidance explains that regulated organizations must assess risks to electronic protected health information and does not prescribe one storage system. Only an authorized record location should hold these data. Applicable access, correction, retention, disclosure, export, backup, and disposal rules still govern the record.

Version and correction log

DateProduct contract or rubric versionOriginal entryCorrected entryReason and source evidenceAuthorizerSummaries affectedPeople notified

Preserve the original entry and an audit trail. Do not backdate a rubric, overwrite attribution uncertainty, or rescore an altered artifact without labeling the change.

Stop and route concerns

Pause the review when the expected set, product definition, attribution evidence, rubric version, review window, authorized record location, client authorization, safety process, or responsible clinician is unclear. Route suspected data alteration, inappropriate surveillance, privacy exposure, inaccessible communication, coercion, acute risk, or clinical uncertainty through the appropriate organizational, clinical, privacy, security, compliance, legal, or emergency path.

A polished artifact is not necessarily valid data. Preserve what is unavailable and unknown, keep the raw evidence, and obtain qualified review before changing care.

Related resources

Sources