An ABA token economy worksheet should show more than a row of empty circles. The useful record explains what earns a token, what the token can be exchanged for, when exchange is available, what happens when an opportunity is unavailable, and whether the system is working as designed for this particular client. Without those details, a polished token board can conceal unclear expectations, inaccessible choices, delayed delivery, or a system that no longer functions as reinforcement.

This copyable worksheet separates design, earning, exchange, client performance, implementation fidelity, and fading questions. It does not choose a target or authorize a token system. The fictional art-studio example makes the arithmetic visible without supplying a clinical threshold, price, schedule, backup item, or recommendation for a real person.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

Clinical boundary: Use only within an authorized, individualized plan and preserve ordinary communication, sensory, mobility, health, safety, and access supports. A token must never become the price of water, food, toileting, communication, safety, education, property, or another basic right. Pause when assent is withdrawn, the system becomes coercive or confusing, an agreed option is unavailable, or distress or another unwanted effect appears. The BACB ethics resources provide professional context; this worksheet cannot establish consent, necessity, effectiveness, payer coverage, or legal compliance.

A token system has connected parts

A token is an item, mark, point, or digital entry that is earned under a defined arrangement and later exchanged for another outcome. The token's appearance does not make it reinforcing. Its function depends on the learning history and current arrangement. A sticker chart without an exchange rule may be a visual record, while a point system tied to exchange may be a token economy. Response cost, in which tokens are removed, is another procedure and should not be smuggled into a positive-earning plan.

A systematic review of procedural descriptions described six central components: target responses, the token, backup reinforcers, and interconnected schedules for producing tokens, producing exchange, and exchanging tokens. Only 18 of 96 reviewed articles fully described all primary components. That finding supports explicit documentation, not a universal six-box recipe.

The current BCBA Test Content Outline includes conditioned reinforcement, schedules, preference assessment, generalization, maintenance, integrity, and data-based evaluation. The BACB outline access page identifies the current outlines. Neither source decides whether a token system is appropriate for a particular client.

Purpose, direct measure, and participation

Begin with a purpose the client understands and values. “Be compliant” is not a sufficient target. Name the setting, opportunity, observable response, accepted variations, measurement unit, ordinary supports, and why the outcome matters. Record how the client can choose, decline, ask questions, request a different option, take a break, or report that the system is no longer useful.

Blank design card

FieldCase-specific entryClient-selected or agreed purposeSetting, routine, people, and meaningful outcomeObservable target response and accepted variationsOpportunity start/end and direct measureOrdinary AAC, sensory, motor, visual, hearing, and processing supportsHow choice, assent, dissent, help, break, and change are communicatedBaseline or other evidence consideredWhy a token arrangement is being consideredAlternatives considered and client feedbackClinical owner, version, effective date, and next review

ASHA's AAC Practice Portal describes AAC as individualized and multimodal. Do not remove a communication device or familiar support to manufacture “independence.” A supported but valid response can remain valid under the case definition even when an additional teaching prompt would change its score.

Token, backup options, and current evidence

Record the exact token form and how a reviewer will distinguish an earned token from a decorative icon, sample, duplicate, or stale digital balance. Identify the evidence that the token has acquired value in this arrangement. A preference assessment can inform backup options, but selecting an item does not prove that a token paired with it will strengthen the target response.

In a small assessment of token reinforcement for two students with autism, the reinforcing value of an established token system varied between participants. The study supports checking function directly rather than assuming that a familiar token board works for everyone.

Design elementRecordToken identity, format, and controlled sourceHow earned, sample, duplicate, lost, or damaged tokens are distinguishedBackup options and date of current preference evidenceWhat can be chosen at exchangeToken price or exchange relation for each optionWhat happens if an option becomes unavailableEvidence that tokens function in this arrangementClient feedback, burden, stigma, or unwanted effects

Use neutral language. “Did not care enough” is not a measurement result. A change in responding might reflect token value, backup value, task difficulty, timing, competing events, access problems, unclear rules, implementation mismatch, or another variable. The record should preserve those possibilities.

Earning and exchange rules

The earning rule says what response produces how many tokens and when. The exchange-production rule says what makes an exchange period available. The exchange rule says what a token balance purchases. Those are separate schedules. State each one in plain language that the client and implementers can understand.

Versioned operating map

ComponentCurrent ruleClient-facing explanationExceptionEvidence/sourceVersion ownerToken-production scheduleExchange-production scheduleToken-to-backup exchange relationDelivery timing windowExchange timing and locationCarryover, saving, reset, loss, or duplicationPrompt/correction interactionStop, hold, and review rule

Do not add token loss simply because an implementer wants a faster consequence. If a qualified plan separately includes response cost, name it clearly instead of hiding it under “token correction.” Review the procedure's risks, alternatives, consent or assent implications, data, and authorization. This public ABA token economy worksheet does not decide that such a procedure is justified.

Earning-opportunity ledger

The event record preserves planned coverage and valid performance. An opportunity may be invalid when materials are unavailable, the definition changed, the implementer gave the token before the response, the client could not access the instruction, or the environment ended the opportunity. Invalid does not mean the client failed.

Date/timeContextTarget/opportunity IDPlanned?Available and accessible?Initial responseMeets case definition?Token earnedDelivered within window?Prompt/correctionValid? reasonAssent/dissent or unwanted effectObserveryes / no / not scorableyes / no / not scorableyes / no / not scorable

Keep the target-response result separate from token-delivery fidelity. If the client produces the response and the implementer delivers the token late, the client result may remain correct while implementation does not. Conversely, correct delivery cannot convert a response that missed the definition into a success.

Exchange-visit record

An earned balance is not useful if exchange is inaccessible, options are unavailable, prices change without notice, or the client cannot communicate a choice. Capture the exchange attempt and outcome directly.

Visit ID/dateBalance presentedExchange available?Options actually availableChoice communicatedPrice appliedExchange completed?DelayDeclined/deferred/changedValid? reasonClient responseImplementer

A declined exchange is not automatically a problem. It may reflect saving, changed preference, inaccessible choices, competing activities, fatigue, or simple noninterest. Record what happened without inventing a motive. Never erase an earned balance merely to force an immediate choice unless a separately reviewed rule lawfully and ethically governs carryover.

Implementation-fidelity record

Fidelity asks whether the team implemented the current version. It is not an outcome measure. A staff-training study using manualized token-economy instruction evaluated training, generalization, and maintenance of staff skills. Its small design does not prescribe staff training here, but it shows why implementer behavior deserves direct measurement.

ObservationCorrect opportunity identifiedTarget scored before consequenceCorrect token amountDelivered within timing ruleBalance accurateExchange rule followedAccess/choice/response recordedCorrect componentsOpportunitiesNotes

Define every component before observing. “Used token system correctly” is too vague to reproduce. A component can be not observable rather than silently scored incorrect. Fidelity mismatches should trigger an implementation question, not an automatic conclusion about client motivation or treatment effect.

Denominators that answer different questions

Review questionNumeratorDenominatorResultWhat it does not showValid opportunity coveragevalid earning opportunitiesplanned earning opportunitiestarget performancePerformance among valid opportunitiesvalid opportunities meeting the target rulevalid earning opportunitiesplanned coverage or causationPlanned-opportunity yieldvalid opportunities meeting the target ruleall planned opportunitieswhy events were invalidTimely token deliveryearned tokens delivered within the defined windowearned tokens duetoken function or client outcomeCompleted valid exchange visitsvalid visits with independent choice and completed exchangevalid exchange visitsfuture preference or reinforcementImplementation fidelitycorrectly implemented componentsobservable component opportunitieseffectiveness

The same number must not serve every question. If 56 of 60 earning opportunities are valid, 44 meet the target definition, and 41 earned tokens arrive on time, the denominators are 60, 56, and 44. Combining them would hide either unavailable opportunities or delivery problems.

Fictional community art-studio example

The fictional client chose a community art-studio setup routine as meaningful. The invented plan defined six earning opportunities per session across 10 sessions. Four planned opportunities were invalid because the relevant station or materials were unavailable, leaving 56 valid opportunities. Forty-four valid opportunities met the case-defined target.

  • valid opportunity coverage: 56 / 60 × 100 = 93.3%;
  • performance among valid opportunities: 44 / 56 × 100 = 78.6%; and
  • planned-opportunity yield: 44 / 60 × 100 = 73.3%.

Forty-four tokens were due. Forty-one were delivered within the fictional timing window, so timely delivery was 41 / 44 × 100 = 93.2%. Twelve exchange visits were planned. One was unavailable because every agreed option was absent. Nine of 11 valid visits included an independent choice and completed exchange, or 9 / 11 × 100 = 81.8%.

The team separately reviewed 20 observations across seven implementation components, creating 140 component opportunities. One hundred twenty-nine were correct and 11 were mismatches, reconciling to 140. Implementation fidelity was 129 / 140 × 100 = 92.1%. That percentage describes observed implementation, not whether the token system caused a change.

Fictional review table

Evidence layerFictional resultBounded interpretationPlanned coverage56 valid of 60 plannedFour events were not interpretable as client performanceTarget performance44 of 56 validDescribes the current definition and contexts onlyPlanned yield44 of 60 plannedRetains implementation availability in viewTimely delivery41 of 44 earnedThree earned tokens were lateValid exchange completion9 of 11 valid visitsOne planned visit was unavailable; one valid visit did not completeImplementation fidelity129 of 140 componentsEleven component mismatches require reviewGeneralization or maintenancenot establishedNo durable or cross-context claim follows

The numbers invite a review of unavailable materials, late deliveries, exchange access, current preferences, unwanted effects, and the exact response pattern. They do not generate an automatic continue, thin, fade, or stop decision.

Fading, thinning, and transfer questions

Reducing token visibility is not the same as thinning token production, changing exchange access, increasing price, transferring control to naturally occurring outcomes, or ending the system. Name the element being changed and preserve the prior version. A systematic review of maintenance strategies found 37 qualifying studies and described thinning, fading, transfer of control, combined approaches, and partial retention. The authors also noted limited direct comparisons, so the review does not identify a universal best sequence.

Change versionQuestionElement changedElements held constantClient feedbackDirect resultFidelityUnwanted effectRollback ruleDecision ownerthinning / fading / transfer / removal

Do not fade the client's ordinary supports or earned rights. A change plan should state what evidence would make the team pause or restore the earlier version. The CASP ABA Practice Guidelines access page supplies current professional context for individualized care; it does not endorse this worksheet or decide a fading plan.

Decision and exception record

FieldReview entryVersions and date range reviewedClient feedback, choice, assent/dissent, burden, stigma, or unwanted effectsTarget-performance pattern and alternative explanationsInvalid opportunities, delivery delays, exchange failures, or fidelity mismatchesToken-function or preference evidence still missingGeneralization, maintenance, or fading evidence still missingRights, privacy, safety, school, workplace, payer, or legal issue for separate reviewHold / revise / continue / assess another option questionQualified decision owner and next review date

The Ethics Code for Behavior Analysts keeps competence, client rights, informed consent, assent when applicable, documentation, confidentiality, and data responsibilities with the responsible professional. A completed form is not evidence that those duties were met. Preserve disagreements, missing information, exceptions, and the accountable reviewer.

Related resources

Sources

The studies and reviews span different participants, settings, targets, token arrangements, staff-training methods, and outcomes. They support direct, component-level documentation and cautious evaluation, not a standard price, schedule, backup option, response-cost rule, fading sequence, or claim that a token system is effective for a particular client.