An ABA single stimulus preference assessment data sheet should preserve what the person actually encountered. A mark beside an item is difficult to interpret if the record omits its representation, access method, observation window, or the difference between approach, engagement, rejection, and no response.

The ABA single stimulus preference assessment data sheet below follows one candidate at a time from preparation through review. It keeps item-specific exposure, engagement time, access delivery, burden, and fidelity separate. Name the decision question before the first presentation. A result describes responding within a defined assessment version; it does not turn an activity into a permanent preference or a demonstrated reinforcer.

Clinicians & ABA Professionals / Assessment and Treatment Planning.

Clinical and access boundary: A qualified team must decide whether a single-stimulus format fits the person, assessment question, candidate set, and setting. Preserve AAC, communication, sensory, motor, visual, hearing, literacy, language, mobility, and other ordinary access. Honor help, pause, break, stop, rejection, distress, pain, fatigue, and withdrawal under the person’s plan. This worksheet does not authorize forced approach, prompted consumption, blocked communication, deprivation, physical guidance, access restriction, or a treatment change.

One option at a time still requires a clear question

In a single-stimulus assessment, a defined option is presented by itself and the assessor observes a prospectively defined response. Depending on the qualified design, the record might emphasize approach, use, consumption, active engagement, passive engagement, rejection, or some combination. Presenting one option can reduce the demands of comparing an array, but it changes the question: the person is responding to that candidate in that presentation, not choosing it over a simultaneously available alternative.

The original Pace and colleagues study presented items singly and measured approach in a small, specific sample. A later single-stimulus engagement study measured engagement duration and found that its rankings predicted relative reinforcer effects for four participants, while also finding less stability than paired-stimulus rankings. These studies establish research examples, not a universal script, cutoff, or client conclusion.

The BCBA Test Content Outline, Sixth Edition includes designing and evaluating preference assessments and interpreting assessment data. Those competencies support deliberate method selection. They do not make this blank worksheet a stand-alone assessment protocol.

Record the candidate and its access conditions

Before any presentation, document how the candidate entered the set. A convenient staff-selected object may be easy to score and irrelevant to the person. Direct nomination, rejection, change-of-mind communication, and information from people who know the client all belong in the preparation record. Record that origin explicitly.

Preparation fieldCase-specific entryClient-informed question and meaningful context.How the person nominated, rejected, or changed candidates.Candidate code, exact item or activity, quality, and version.Representation used: object, activity sample, image, video, text, sound, or other.Ordinary access that remains available throughout.Accepted communication, approach, rejection, help, and stop forms.Familiarization and prior contact with each candidate.Observation and access window selected by the qualified owner.Health, allergy, swallowing, sensory, motor, privacy, and safety review.Consent requirements and assent or dissent process.Qualified owner, collaborators, effective date, and review date.

The ASHA AAC Practice Portal describes direct and indirect selection techniques, including partner-assisted scanning. ASHA’s communication-access guidance also centers participation and communication preferences. A familiar device, eye-gaze access, adapted reach, extra processing time, or a reliable partner-assisted signal is ordinary access, not noise to remove from the assessment.

Define what each observation state means

“Interested” is not a reproducible data code. Define the observable boundary for each state before collecting data and distinguish what the person did from what the observer inferred.

StateObservable definitionEvidence capturedPause, invalid, or stop ruleCandidate available as designed.Approach or selection response.Active engagement.Passive engagement or orientation, if clinically justified.Rejection or move-away response.No response or unclear response.Help or access-repair request.Break, decline, or stop.Arranged access delivered.Valid presentation.Invalid presentation.

A study of active and passive approach responses used individualized observable responses with three children who had profound multiple disabilities and minimal movement. That work illustrates why access and response definitions can matter. Keep those observations distinct. The study does not mean looking, smiling, stillness, or any other signal has the same meaning for every person.

Likewise, contact and engagement are not interchangeable. A recent competing-stimulus case summary discusses engagement as a more informative single-stimulus measure than contact for its purpose. A hand resting on an item may be contact without active use; a person may also engage through listening, watching, or directing another person. Write the case-specific definition instead of borrowing a topography uncritically.

Plan presentations without manufacturing differentiation

The form can document a randomized, counterbalanced, or otherwise justified order. The blank form cannot choose the order or number of repetitions. Keep the rationale beside the order. Record planned exposure so invalid or missed presentations cannot silently change denominators.

PresentationCandidate/versionPlanned contextRepresentation and access methodObservation windowCompleted?Valid? reason123Continue for the approved plan.

Do not prompt approach merely to fill a cell. A prompt changes the response being measured. If demonstration or familiarization is needed, label it separately from scored presentations. A comparison of single- and paired-stimulus methods found that both methods identified preferred stimuli in its sample, while paired presentations produced more differentiated hierarchies. That bounded result is a reminder that a ceiling pattern, such as approach to everything, can be an interpretable limitation rather than a reason to create artificial rejection.

Single-stimulus presentation ledger

Use one row for one planned presentation. Preserve the event even when it is invalid or the person does not approach. Keep invalid rows visible.

Date/timePresentation/versionCandidateAccess setupValid? reasonApproachActive engagement secondsPassive engagement or orientationRejectionNo/unclear responseAccess deliveredLatencyHelp/decline/stopBurden or feedbackObserveryes / noyes / noyes / no / not applicable

If the person approaches and then rejects an option, retain both observations. If access does not begin as arranged, record the mismatch rather than coding short engagement as low preference. Missing AAC, a changed picture, an inaccessible position, an unavailable activity, staff steering, or an unplanned delay can invalidate a presentation under the prospective rule.

Keep four denominators visible

Single-stimulus data can look simple while mixing incompatible measures. Reconcile planned presentations, valid item opportunities, valid observation seconds, and access-delivery opportunities separately. Show each denominator beside its numerator.

Candidate/versionPlanned presentationsInvalidValid opportunitiesApproachesApproach percentageValid observation secondsActive engagement secondsEngagement percentageClear rejectionNo/unclear responseAccess mismatchABContinue.

Review measureNumeratorDenominatorWhat it can describeValid coveragevalid presentationsplanned presentationsUsable exposure under the written rulesApproachapproaches to one candidatevalid presentations of that candidateDefined approach in this version and contextEngagementactive-engagement secondsvalid observation seconds for that candidateTime allocation while the option was singly availableAccess deliveryarranged access deliveredapproaches requiring arranged accessImplementation, not preference strengthFidelitycorrect componentsobservable component opportunitiesCorrespondence with the written design

An overall approach rate can hide that one option had fewer valid presentations. An engagement percentage can also answer a different question from approach: a person might approach quickly and engage briefly, or approach less often and remain engaged longer. Neither measure independently proves dislike, benefit, reinforcement, or what will happen in another setting.

Fictional makerspace example

Marisol is a fictional adult exploring six accessible makerspace activities, coded A through F. Marisol nominated four activities directly and agreed to sample two suggested by a familiar support person. A speech-generating device, adapted switch, ordinary headphones, water, and a break area remained available. The example exists to test the worksheet’s arithmetic, not to recommend a candidate set or observation duration.

The team planned four presentations per candidate: 6 x 4 = 24 planned presentations. Three presentations were invalid: one each for D, E, and F. The reasons were a changed representation, an unavailable adapted control, and a staff comment before the response window. That left 24 - 3 = 21 valid presentations, so valid coverage was 21 / 24 x 100 = 87.5%.

CandidateValid opportunitiesApproachesApproach percentageValid observation secondsEngagement secondsEngagement percentageA444 / 4 x 100 = 100.0%480360360 / 480 x 100 = 75.0%B433 / 4 x 100 = 75.0%480300300 / 480 x 100 = 62.5%C422 / 4 x 100 = 50.0%480210210 / 480 x 100 = 43.8%D322 / 3 x 100 = 66.7%360240240 / 360 x 100 = 66.7%E311 / 3 x 100 = 33.3%360120120 / 360 x 100 = 33.3%F311 / 3 x 100 = 33.3%3606060 / 360 x 100 = 16.7%Total2113Item-specific denominators above2,5201,2901,290 / 2,520 x 100 = 51.2%

The opportunity count reconciles as 4 + 4 + 4 + 3 + 3 + 3 = 21. The approaches reconcile as 4 + 3 + 2 + 2 + 1 + 1 = 13. Five other valid presentations included the predefined rejection response, and three had no or unclear response. Those eight observations remain direct data; they are not combined into a “disliked” category.

Every valid presentation used a 120-second observation window, producing 21 x 120 = 2,520 valid observation seconds. Active engagement totaled 360 + 300 + 210 + 240 + 120 + 60 = 1,290 seconds. The overall 51.2% is context, not a score to maximize. Item D illustrates why measures stay separate: its approach percentage was 66.7%, while its engagement share of valid availability was also 66.7% by coincidence, not by mathematical necessity.

Arranged activity access followed 12 of 13 approaches, or 12 / 13 x 100 = 92.3%. One access mismatch prompts a validity review because it may influence later responding. Review the access failure before interpreting later presentations. That mismatch is not evidence about Marisol’s preference.

Fidelity and interpretation review

ObservationCorrect versionCandidate exactAccess availableResponse window honoredNo steeringApproach coded accuratelyEngagement timed accuratelyRejection/no response preservedHelp/stop honoredAccess deliveredBurden recordedCorrectOpportunities

The fictional fidelity sample covered 12 observations across 11 components: 12 x 11 = 132 component opportunities. Reviewers marked 124 correct and eight mismatches, and 124 + 8 = 132. Descriptive fidelity was 124 / 132 x 100 = 93.9%.

Read the mismatch details before the percentage. Inspect the mismatch itself. Failure to honor a stop response or provide the agreed access method can affect welfare and validity even when the aggregate looks high. Fidelity shows whether the written arrangement occurred. Even so, fidelity does not prove the arrangement was appropriate.

Interpretation questionDirect evidencePlausible alternative or limitationFollow-up neededQualified ownerWere the candidates meaningful and distinguishable?Did approach and engagement tell the same story?Did representation or access affect responding?Were rejection, no response, help, and withdrawal honored?Did health, fatigue, sensory, motor, or context factors change?Does the result generalize beyond this setting and version?Is a separate reinforcer-effect question warranted?

The SPADS decision model describes several empirically derived preference-assessment formats and considerations for choosing among them. It is a decision-support model, not a substitute for case-specific competence or client participation. The current CASP and APBA assessment-guideline page similarly frames assessment resources as educational and emphasizes individual circumstances and independent professional judgment.

Decision and version record

Review fieldEntryQuestion, candidate set, representations, and procedure version.Client nomination, direct feedback, rejection, help, and withdrawal.Valid opportunities, approach, engagement, and access results.Fidelity mismatches, burden, and validity limits.What the data support and do not support.Retain / revise candidate / change format / pause / stop question.Qualified owner, rationale, effective date, and next review.

Keep raw rows and superseded versions. Version changes should remain visible. A different representation, communication method, observation window, activity quality, access arrangement, setting, or candidate set creates changed conditions. The BACB Ethics Code emphasizes benefit, dignity, self-determination, competence, consent, assent when applicable, and accurate documentation. The code does not endorse this worksheet or replace applicable law, licensure, organizational policy, payer terms, or interdisciplinary responsibilities.

Related resources

Sources

These studies use small samples, particular candidates, operational definitions, settings, and access arrangements. They support direct measurement and qualified interpretation. They do not establish one best response definition, presentation length, repetition count, hierarchy, cutoff, reinforcer, treatment, or expected result.