Response class

A class groups individual responses

A response is one occurrence. A response class groups multiple occurrences according to properties relevant to the analysis. The boundary is selected for a purpose and tested against data; it is not a natural label that permanently belongs to a person.

A single response can fit several descriptive classes. Saying “stop” once can belong to a class of spoken words, one-word messages, or requests to end an activity. The useful class depends on the question. Counting all one-word messages answers a different question from counting every accessible way the person asks to stop.

The Association for Behavior Analysis International basic-principles page lists response classes and functional relations within higher-education content. The BACB BCBA Test Content Outline, 6th edition covers distinguishing behavior, response, and response class in B.1, operational definitions in C.1, representative and reliable measurement in C.8 and C.9, and experimental design in Domain D. These sources identify training and examination content, rather than a class for one client.

Topographical classes group shared form

A topographical response class groups instances with specified physical features. All presses of a large switch using a hand movement could form one class when the definition states the required contact, force, and completion boundary. Clapping may form another class when both hands make contact and separate according to a stated rule.

Topography can support consistent observation, yet physical similarity alone does not show why responses occur. Two identical button presses may participate in different relations if one turns on music and the other requests assistance. Broad labels such as “aggression,” “disruption,” or “noncompliance” can also collapse unlike events and obscure pain, access barriers, communication, and ordinary refusal.

A sound definition says what counts, what falls outside the class, where an instance begins and ends, and how separated occurrences are scored. It also identifies context, ordinary supports, observer, measure, and period.

Functional classes require evidence of common effect

In common applied usage, a functional response class contains different response forms that produce and are maintained by the same consequence under defined conditions. A person may request a pause through speech, an AAC selection, a break card, or an agreed gesture. If those forms reliably produce a pause and their occurrence varies with relevant conditions, treating them as one functional class may help answer a clinical question.

One shared outcome is not enough. Several responses can be followed by the same event by coincidence, a schedule, or partner habit. Interviews and descriptive observations can suggest a functional class. Strong causal claims require appropriate experimental evidence selected by a qualified clinician, with health, safety, feasibility, and assent considered.

Terminology varies. Palmer's conceptual article proposes reserving response class for units that covary partly through shared topography and using functional class for different forms united by common consequences. Many applied sources use functional response class for the latter. Writers should state their convention, membership rule, and evidence rather than assume every reader uses the same boundary.

Class members may not occur equally

Responses in one functional class can differ in effort, speed, history, accessibility, and how reliably partners respond. A person may use the fastest effective form first and another form when the first goes unrecognized. Changes to one class member can shift responding to another, so monitoring only one form can produce a misleading trend.

In a translational study with 11 college students, Mendres and Borrero arranged three clicking responses that produced points or avoided point losses. Participants usually selected the least effortful response available within the arranged class. This analogue study supports attention to response effort and allocation in that laboratory task. It does not supply a universal hierarchy or clinical rate.

For communication, preserve form-specific data alongside the class total. A stable total can hide lost AAC access; a falling speech count can coexist with effective gesture or device use. Partner recognition, wait time, device readiness, and whether the requested outcome was available deserve their own measures.

A fictional example makes the two levels visible

Avery is a fictional sixteen-year-old who chooses a goal of requesting a pause during a community cooking class through speech, AAC, or a break card. An eligible opportunity begins when Avery is participating, a pause is available, all three communication forms are accessible, and the class has not ended. One request is scored per opportunity before any prompt.

Across ten eligible opportunities, Avery requests a pause in 7 of 10. The seven responses include three AAC selections, two spoken requests, and two card exchanges. Partners acknowledge the request within 20 seconds in 6 of 7. The missed acknowledgment remains in the partner denominator.

The class total is seven responses, while the form counts are 3, 2, and 2. These data describe one observation period. They cannot establish that all forms belong to one functional class, that the plan caused responding, or that a higher request rate is better. The team checks Avery's report, pause quality, task fit, prompt use, access, and partner accuracy before deciding what to change.

Good class definitions protect communication and choice

Include every reliable response form that matters to the question. Do not require speech, eye contact, a particular motor action, or device use when another accessible message is clear. Keep food needed for adequate nutrition, water, bathroom access, communication and AAC, mobility, pain care, prescribed treatment, safety, and emergency help outside performance contingencies.

For BCBA and BCaBA certificants and people who have completed an application for either credential, the current BACB Ethics Code addresses competence, client involvement, understandable communication, informed consent and assent when applicable, medical needs, assessment, data, and continual evaluation. The BACB has no separate jurisdiction over organizations or corporations.

Revisit class membership when health, setting, consequence availability, communication, partner behavior, or response form changes. A class is useful when it improves the accuracy and humanity of the analysis. It becomes harmful when it erases the person's message or turns different events into one convenient label.