Brief functional analysis

Safety status: Before any experimental analysis, use a documented risk assessment, medical and interdisciplinary review when indicated, consent and assent when applicable, trained staff, stop rules, protective and emergency plans, and the least risky method capable of answering the question.

A brief FA remains an experiment

Interviews, rating scales, ABC observations, record review, and scatterplots may suggest hypotheses. They involve no planned manipulation of a suspected variable. A functional analysis changes environmental conditions and compares measured responding across them.

“Brief” describes an abbreviated format, not a lower evidentiary standard. A credible comparison, clear definitions, treatment integrity, and data still matter.

The original model used rapid comparisons

Northup and colleagues reported a 1991 outpatient evaluation of three people in which test conditions for aggressive behavior and a contingency-reversal phase involving an alternative communication response fit within a 90-minute visit. The study was small, highly structured, and specific to those cases.

Later applications vary in session length, design, sequence, responses measured, and decision rules. Copying “90 minutes” or another published feature cannot create a valid brief FA. The clinician must select a design that fits the question and case.

Efficiency changes the evidence available

A shorter analysis can reduce assessment burden and help in outpatient, school, consultation, or other time-limited settings. It also yields fewer observations and may miss low-rate, delayed, variable, or context-specific patterns.

A review of FA efficiency included 15 brief FAs from five research articles under its criteria. The authors emphasized design-specific evidence and limitations rather than a universal duration or success rate. An undifferentiated or inconsistent brief result can support further assessment instead of a forced conclusion.

Select the method after preliminary assessment

Review records, health variables, communication, client priorities, operational definitions, history, direct observations, suspected variables, and feasible control conditions. Ask whether another method can answer the question with less risk.

A brief FA may fit when the target response occurs often enough for a short comparison, conditions can be arranged with integrity, and risk can be managed. Low-frequency events, serious injury risk, unclear measurement, unavailable medical information, or weak environmental control may support another route.

Safety is a design feature

Functional analysis may arrange conditions in which harmful behavior becomes more likely. The functional-analysis risk-assessment paper offers a decision-tool framework for considering risk; it is a scholarly proposal rather than a legal safe harbor or universal checklist.

Define entry and stop criteria, session and cumulative exposure limits, staffing, protective equipment when clinically indicated, precursor or latency measures when supported, communication access, medical escalation, and post-session review. Avoid exposing a person to danger merely to obtain clearer data.

Emergency action, mandated reporting, and needed medical care bypass routine assessment steps. AAC, food, water, bathroom access, mobility, prescribed care, and emergency help stay available.

Interpretation needs comparison and replication

Evaluate level, trend, overlap, immediacy, replication, integrity, exposure, and plausible alternative explanations. One elevated test segment can reflect sequence effects, carryover, fatigue, novelty, or implementation error.

Rate, frequency, latency, duration, or another measure should fit the response. Use the same measurement rules across compared conditions. Report excluded segments and stop-rule activations separately.

A fictional decision example

Jae is a fictional client whose qualified team considers a brief FA after medical review, interviews, and observation. Before assessment, the team defines 12 planned segments across selected test and control conditions, along with fixed safety stops. All 12 segments reach a valid observation endpoint.

Eight segments meet the integrity criterion, so valid-integrity coverage is 8 of 12, or 66.7%. The target response occurs in three test segments and one control segment. Those counts alone provide no functional conclusion: conditions have different roles, the sample is small, and the analysis needs visual and design-based interpretation by the qualified clinician.

One additional segment stopped for a safety threshold would remain visible in the planned-exposure record. Removing it from every denominator would hide assessment burden.

Keep treatment decisions separate

An FA result informs treatment selection. It never automatically prescribes a goal, dosage, procedure, or setting. The clinician should integrate results with client priorities, risks, communication, preferences, feasibility, other assessment evidence, and interdisciplinary findings.

If results are unclear, record “inconclusive” and choose the next evidence step. Options may include improved definitions, further observation, another FA design, a different comparison, medical referral, or stopping the experimental path.

The BACB Test Content Outline covers assessment and functional analysis as examination content. The current BACB Ethics Code addresses competence, medical needs, consent and assent when applicable, assessment, intervention, risk, data, documentation, and evaluation for covered professionals.