Functional analysis

An FA tests a defined relation

A clinician first defines the response, the environmental variables to examine, and the decision the assessment should support. The clinician then arranges at least two conditions. A test condition changes a selected antecedent or consequence. A comparison condition holds relevant features differently. Direct measurement shows whether responding changes reliably with that contrast.

The ABAI Basic Principles in Behavior Analysis page places functional analysis within higher-education content on experimental analysis and behavior-environment relations. It is a training resource, rather than a clinical authorization or step-by-step protocol. The current BACB BCBA Test Content Outline, 6th edition separately covers descriptive assessment in F.5 and designing and evaluating functional analyses in F.6. The outline describes examination content and does not establish who may practice in a state or setting.

The classic Iwata and colleagues study repeatedly exposed nine participants to analogue conditions that differed in demands, attention, and materials. Six showed consistently higher self-injury in a specific condition. That foundational study illustrates experimental comparison; it is not a template that should be copied for every person, response, or setting.

Functional analysis and FBA differ

An interview, rating scale, ABC record, or scatterplot does not become an FA because it concerns function. An FA requires planned manipulation and a credible comparison. An FA can also produce an undifferentiated or inconclusive result. That result should remain inconclusive while the team examines measurement, condition fidelity, relevant variables, context, and whether another assessment route is warranted.

Hanley, Iwata, and McCord defined review inclusion around direct observation of problem behavior under at least two conditions with an environmental manipulation intended to demonstrate a relation. A later 40-year review reported 1,333 FA outcomes from 326 studies published from June 2012 through May 2022. The literature contains varied participants, conditions, settings, durations, and outcomes. Its size does not make one format universally suitable.

Design follows the person and the risk

Before an FA, a qualified clinician should document:

1. the client-informed question and why experimental comparison adds value
2. the response, measurement system, eligible opportunities, and comparison logic
3. consent, assent when applicable, and accessible explanations
4. medical, trauma, sensory, communication, environmental, and interdisciplinary considerations
5. ordinary supports that remain available in every condition
6. risks, safeguards, trained roles, stop rules, emergency response, and setting
7. procedural-integrity and observer-agreement checks
8. how results, limits, and next decisions will be shared

Some questions can be tested through briefer components, latency, a lower-risk precursor, or another response. Sometimes experimental analysis should be deferred or declined. The choice depends on the person, the response, available safeguards, and whether the expected information justifies the exposure. A clinician should never evoke dangerous behavior simply to complete a standard condition sequence.

For covered BCBA and BCaBA certificants and applicants, the current BACB Ethics Code addresses understandable communication, client and stakeholder involvement, informed consent and assent when applicable, medical needs, scientifically supported assessment, contextual fit, risk minimization, written results, and appropriate data use. The BACB has no separate jurisdiction over organizations or corporations. Licensure, payer, facility, school, employment, and other rules still need their own review.

Communication and essential access stay available

Speech is one communication form. A response definition can include augmentative and alternative communication, or AAC, gesture, sign, writing, or another reliable form. ASHA's AAC practice portal says people who use AAC should always have access to their tools or devices. The team should confirm positioning, charging, vocabulary, motor access, partner recognition, and a backup method before interpreting data.

Food and water, bathroom access, mobility, prescribed care, pain care, communication, and immediate safety should never be withheld as experimental variables. A recognizable withdrawal of assent, distress signal, or medical concern activates the defined stop and care pathway. The record should retain stopped and excluded components with their reasons.

A fictional low-risk example

Lee is a fictional twelve-year-old who wants adults to respond consistently to a spoken or AAC pause message during a chosen puzzle activity. With Lee, a qualified clinician defines an eligible opportunity as a two-minute puzzle period with AAC present, no sign of distress, and freedom to end participation at any time. A pause message is counted before a prompt. Staff honor any distress or withdrawal immediately, outside the experimental comparison.

The clinician alternates two preplanned conditions across three matched pairs. In the contingent-pause condition, a recognized message produces a 30-second pause. In the comparison condition, matched pauses occur on a time-based schedule independent of Lee's message; recognized messages still receive acknowledgment and never delay a requested end. Lee gives 12 pause messages across 18 eligible minutes in the contingent condition and 3 across 18 eligible minutes in the comparison condition. The separation appears in each of the three pairs.

This fictional pattern supports a functional relation between the contingent pause and Lee's message within those narrow conditions. It does not show why every message occurs elsewhere or establish a treatment goal. The report includes raw counts, exposure time, condition order, supports, integrity, agreement, assent withdrawals, exclusions, and Lee's account. Planning could focus on reliable partner response and broader access to the pause message because the message is useful communication.