Understanding FBA and function-based behavior support

Behavior is information. When a student leaves class, refuses work, calls out, shuts down, or destroys materials, the useful question is not how to stop it. It is what conditions predict it, what follows it, and what the behavior is currently accomplishing. This is a working reference on how that question gets answered and turned into a plan.

Define→Collect→Find the pattern →Build the pathway→Teach a replacement →Implement→Measure outcome and fidelity →Revise

What an FBA is

A functional behavioral assessment is a systematic process for understanding the function of a specific interfering behavior, and the factors contributing to when it occurs and when it does not. Its purpose is to build positive, function-based supports rather than relying on reactive or exclusionary responses.

An FBA is not a test score. It is an investigation assembled from converging data, and it produces a hypothesis rather than a finding. That word matters: an FBA does not diagnose, establish intent, or determine whether a behavior was willful. It produces the best current explanation of what a behavior accomplishes, which the plan is then built to test.

One point of contemporary federal guidance deserves emphasis, because it contradicts how FBA is often treated in practice: an FBA can be considered for any student whose behavior interferes with learning — with or without a disability, and for internalizing behavior as well as externalizing behavior. It is not only a compliance procedure used after a disciplinary event — IDEA does attach specific FBA obligations in certain disciplinary circumstances, covered below. But treating it as a post-discipline artifact means the process arrives long after the point at which it would have helped.

Function, not form

Topography — what the behavior looks like — matters a great deal for safety, measurement, severity, and the logistics of responding. What it cannot do on its own is tell you which intervention will work.

Two students throw materials during independent work. The behavior is topographically identical. If the team responds to the topography, both get the same plan. If the team responds to the function, they get opposite plans, because the behavior is doing opposite things.

Swap the plans and both fail. Student A gets scheduled adult attention, which does nothing about a worksheet he cannot read, and his throwing still ends in the task being removed. Student B gets a break card — a route to less adult contact, which is the opposite of what his behavior is producing. Neither plan is poorly written. Each is matched to the wrong contingency.

Define it before you measure it

Words like noncompliant, disrespectful, disruptive, unmotivated, and aggressive are too subjective to measure consistently. Two people watching the same ten minutes will not produce the same count, which means nothing built on that count can be trusted.

An operational definition describes what the behavior actually looks and sounds like. A useful one is observable (another person can see or hear it), measurable (it can be counted or timed reliably), objective (it describes behavior rather than motives, emotions, or character), and bounded (observers know when an episode starts and stops).

The working standard: an unfamiliar observer should be able to walk into the room and determine whether the behavior just occurred.

Not measurable	Measurable
“Is defiant.”	Does not begin the assigned task within 60 seconds of a direction, following one restatement.
“Is disruptive.”	Speaks at conversational volume or louder without being called on, during teacher-led instruction.
“Shuts down.”	Places head on desk and does not respond to two consecutive prompts, for 60 seconds or longer.
“Is aggressive.”	Makes forceful physical contact with another person using a hand, foot, or object.

A complete definition also carries examples and non-examples. The non-examples do the quiet work — they are what stop an observer from counting an accidental bump, or a raised hand that arrived a second late.

Measure the right dimension

Different behaviors call for different measures, and choosing the wrong one produces data that cannot show change even when change is happening.

Frequency: How many times. Use for discrete behaviors of similar length within a consistent observation window.
Rate: How many times per unit of time. Use when observation periods vary in length, which in schools they almost always do.
Duration: How long. Use when the concern is that episodes last too long rather than that they happen too often.
Latency: How long between a direction and the start of the behavior. Often the right measure for task initiation and refusal.
Interval: Whether the behavior occurred during each of a series of short intervals. An estimate, used when continuous recording is not feasible.

These answer how much. ABC observation answers a different question entirely — what was happening around it — and a strong school-based FBA generally needs both contextual and quantitative information. Frequency data with no context cannot generate a hypothesis. ABC data with no quantitative baseline cannot show whether anything changed.

Setting events, antecedents, precursors

These three get used interchangeably and they occupy different positions in time, which means they call for different responses.

A setting event is distinguished from the immediate antecedent because it alters the likelihood or value of later events rather than occurring immediately before the response — the same worksheet is a different event after four hours of sleep than after nine. The precursor is a distinct third thing, and the distinction is practical: setting events are addressed by changing conditions, antecedents by changing the immediate environment, precursors by responding earlier.

Precursor assessment is a real but small corner of the research literature — roughly two percent of functional analyses in a recent forty-year review. Treat a candidate precursor as a hypothesis in its own right until the temporal relationship has actually been observed, not because someone reports that it usually happens first.

Where ABC data goes wrong

The antecedent is too far back. “Had a bad morning” is a setting event. The antecedent is what happened in the seconds before.
Only the loud instances get recorded. Data collected when someone remembers to collect it oversamples severe episodes and hides the pattern.
Non-occurrence is never recorded. When the behavior does not happen is where the plan’s antecedent strategies come from.
The consequence column records what the adult intended. Record what the student actually received. Intent is not function — and a consequence is simply what happened next, not automatically a reinforcer. Whether it strengthened the behavior is the question the assessment is trying to answer, not an assumption to start from.
One sequence becomes a conclusion. A single ABC entry is an anecdote. Patterns require repetition across occasions, settings, and observers.

How functions are organized

Many educators learned this as SEAT. The mnemonic still works, and contemporary functional-assessment language has made two of its four terms more precise.

S — Sensory: Now usually described as automatic reinforcement: the behavior produces its own consequence and needs no other person.
E — Escape: Escape from or avoidance of a task, demand, interaction, or transition.
A — Attention: Attention from an adult or a peer.
T — Tangible: Often broadened to access to items or activities.

Those four come from two underlying questions — is the student getting something or avoiding something, and does another person have to be involved for it to happen?

Seeing the questions rather than the list matters, because real cases do not always land in one box. Contemporary functional analysis research describes multiple and combined functions more often than a four-item mnemonic suggests.

Both cells in the bottom row are automatic reinforcement — the S in SEAT. They differ only in whether the behavior produces stimulation or reduces something aversive, and neither requires another person to respond. These are organizing constructs, not personality types. A behavior can serve more than one function, and the same topography can serve different functions in different settings — escape in reading, attention at lunch. A single hypothesis covering every context is usually a sign the assessment was too narrow.

Diagnosis is context, not function. “He hits because he is autistic” and “she avoids work because she has ADHD” are not functional explanations. They name a category and stop.

Diagnosis tells a team which variables are likely to matter — communication, executive function, sensory, medical, developmental, academic skill. It does not tell them what this behavior accomplishes for this student in this room. For a student with ADHD the data may implicate task length, delayed reinforcement, or movement. For an autistic student, predictability, communication, or sensory conditions. For a student with an intellectual disability, task difficulty and communication repertoire. For a student with no diagnosis at all, the identical logic applies and produces an identical quality of answer.

Converging evidence

For a school-based FBA, no single indirect or descriptive source should be treated as establishing function. Teams strengthen a hypothesis by examining convergence across interviews, records, direct observation, and quantitative data. Experimental functional analysis is different in kind: systematic manipulation of conditions can test a causal relation directly.

The QABF is a 25-item indirect functional-assessment rating scale with five subscales: Attention, Escape, Tangible, Physical, and Nonsocial. Elevated scores contribute evidence toward a functional hypothesis; they do not independently establish function. Its psychometric base comes largely from research in populations with intellectual and developmental disabilities, which is worth knowing before applying it across a general school caseload, and the literature recommends pairing rating scales with observational or experimental methods.

The confidence of the conclusion has to match the method that produced it. An experimental functional analysis, in which conditions are systematically arranged and withdrawn, can demonstrate a functional relation rather than relying on an inferred one. It remains the strongest available method. Descriptive assessment — which is what most school FBAs consist of — supports a hypothesis.

The fix costs nothing. “Observational data across four settings are consistent with an escape-related function” is more defensible than “the behavior is escape-maintained,” and it leaves room for the plan itself to serve as the test. Repeated observations showing a consistent relationship with math demands are worth acting on provisionally. They are not proof of function.

The student and family are data sources

They are partners in the assessment, not just recipients of the plan. Their perspective can surface context, history, preferences, feasible replacement behaviors, and consequences that are not visible during a classroom observation. Student involvement matters more as students get older, and at the secondary level a plan built without the student’s input is usually a plan the student has no reason to cooperate with.

Look for missing skills, not just motivation

Hypothesis development should examine whether the student has the academic, communication, executive-function, or social skill the situation requires. This is where most of the actionable information lives.

“Escape from writing” is thin. “Task exceeds current written-expression skills, no efficient way to request help is available, refusal reliably results in the task being withdrawn” names three things a plan can change. The first version leads to a reinforcement system. The second leads to instruction.

From hypothesis to plan

The competing behavior pathway is the standard bridge. It takes the hypothesis and lays it alongside two alternatives: what the student would ideally do, and what the student can be taught to do right now instead.

The replacement must be functionally equivalent. If screaming produces escape from a difficult task, “sit quietly” is not equivalent — requesting help or a brief break might be. If aggression produces peer attention, independently completing the worksheet does not produce the same outcome.

A workable replacement is functionally equivalent, efficient enough to compete, feasible in the setting, and appropriate for the student’s age and context. Reinforcement is typically dense at first and thinned deliberately as the skill establishes. Schedule design is individualized, but the general principle holds: a replacement that works inconsistently is competing against a target behavior with a long history of contacting reinforcement.

The desired behavior stays on the diagram for a reason. It is the long-term goal and it is usually not achievable yet. Confusing the replacement with the desired behavior is how plans end up asking a student to do the exact thing the assessment just established he cannot currently do.

Prevent, teach, reinforce, respond, monitor

IDEA does not prescribe a BIP format or a mandatory component checklist — that is left to the IEP team, and Arizona guidance states this explicitly. What federal guidance does describe is a recognizable set of elements found in strong plans, and they organize cleanly into five verbs.

The through-line is that a plan should change the environment as well as the student.

Prevent. Change the predictable antecedents and setting conditions so the behavior is less likely to be needed. This is the thinnest section in most plans and the only one that works before anything goes wrong.
Teach. Explicitly instruct the replacement behavior or the missing skill. Named is not taught — specify who teaches it, when, and how mastery will be recognized.
Reinforce. Make the replacement effective enough to compete. Dense at first, thinned deliberately later, never thinned because the adult got busy.
Respond. Specify what adults do when the target behavior occurs, written so it does not deliver the maintaining consequence.
Plan for safety, where risk warrants it. For behavior involving significant risk, keep the proactive function-based plan distinct from the procedures needed to maintain safety during escalation. They serve different purposes and get evaluated differently.
Monitor. Track student outcome and implementation fidelity, with a review date and a decision rule set before the plan starts.

The plan also has to say who is responsible for each piece. A plan that assigns everything to “staff” has assigned it to nobody.

Fidelity is the missing half

A beautifully written plan does nothing if nobody implements it. This is not a philosophical point about behavior work — it is a documented weakness in the actual documents schools produce.

0.08

A 2024 study examined 135 completed FBAs and 129 BIPs from 13 school districts against a technical-adequacy rubric. The documents contained roughly 40–50% of the essential components on average. The mean score for having an implementation fidelity plan was 0.08 on a 0–2 scale. The weakest areas overall were operational definitions, hypothesis-linked intervention, progress monitoring, and fidelity planning — which is to say, the four things this page spends the most time on.

Which means the most common answer to “how will we know this plan was actually delivered?” is that nobody wrote one down. The Center on PBIS has since released the Technical Adequacy Tool for Evaluation as a resource for auditing completed FBAs and BIPs at the district level.

Fidelity matters more in behavior work than almost anywhere else, because a plan is implemented by many adults across many settings. Inconsistent implementation does two things: it obscures whether the intervention is working, and it can allow the target behavior to keep contacting reinforcement.

A single overall fidelity percentage can also hide the variation that matters. A plan may show eighty percent adherence overall while one critical component, one setting, or one adult’s response is implemented inconsistently. Fidelity data are most useful when they show which components were implemented, where, and by whom.

The practical rule is the one that governs all outcome data. Poor outcomes under poor fidelity cannot confidently be attributed to the intervention.

So a useful monitoring system tracks three things, not one: target behavior, replacement behavior, and implementation fidelity.

Target behavior falling is not sufficient on its own. When a plan teaches a defined replacement skill, teams should monitor whether that skill is actually being acquired and used — a reduction in the target behavior does not by itself show the alternative was learned. Fidelity sits on its own panel because it is a percentage and the behavior counts are not. High fidelity with a poor outcome means question the plan. Low fidelity with a poor outcome means establish delivery first, before concluding anything.

PBIS: support is a system

Positive Behavioral Interventions and Supports is a framework that makes individual behavior work sustainable. At Tier 1, expectations are explicitly taught and reinforced for everyone and schoolwide data are used. At Tier 2, targeted supports such as Check-In/Check-Out are layered on. At Tier 3, supports may include individualized, function-based assessment and intervention.

Function-based thinking runs through the whole continuum — Tier 1 and Tier 2 practices are built on the same logic — but formal individualized FBA and BIP work usually appears at Tier 3. Its effectiveness depends heavily on what is happening below it. An escape hypothesis is difficult to act on in a classroom without differentiated instruction. A plan built on dense reinforcement is difficult to sustain where no acknowledgment system exists.

Which means a school producing many individual plans for similar behaviors in the same setting has a systems finding, not a set of individual student findings. Ten escape plans in one classroom is information about instruction.

PBIS is also not an elementary framework. Where high school implementation occurs with fidelity, it has been associated with reductions in discipline referrals and suspensions and improvements in attendance, climate, and perceived safety. And fidelity is measured at the system level too — the Tiered Fidelity Inventory assesses implementation across all three tiers, with recent revisions placing more weight on mental health integration, equity, and classroom implementation.

Discipline, equity, and IDEA

Current federal guidance connects wider use of FBAs and function-based plans directly to building supportive learning environments and reducing unnecessary reliance on exclusionary discipline. It also addresses disparities in disciplinary practice, and warns that implicit bias can affect how educators perceive and classify behavior in the first place.

That warning lands earlier in the process than it might appear. Bias operates at the definition stage — on whether a behavior is recorded as “disrespectful” or as a specific observable action, and on which students get an assessment while others get a referral. Operational definitions are a technical practice and an equity practice at the same time.

What IDEA actually requires

IDEA does not require an FBA every time a student with an IEP has difficulty. It requires that when behavior impedes the student’s learning or that of others, the IEP team must consider positive behavioral interventions and supports and other strategies to address it.

The requirement becomes explicit in one disciplinary circumstance. When conduct resulting in a disciplinary change of placement is determined to be a manifestation of the student’s disability, the team must conduct an FBA if one has not already been conducted and implement a behavior intervention plan — or review and modify an existing plan.

A separate provision applies when the behavior is determined not to be a manifestation, and to removals to an interim alternative educational setting under the special-circumstances provision. In those situations the child receives, as appropriate, a functional behavioral assessment and behavioral intervention services and modifications designed to address the behavior so that it does not recur.

Two things follow, and the first is a distinction worth holding onto. Behavior that impedes learning triggers an obligation to consider positive behavioral interventions and supports. It does not automatically require an FBA at that point. The FBA obligations attach in the disciplinary circumstances above. Conducting one earlier is often good practice; it is a different question from what the statute compels.

Second, where IDEA requires implementation of a behavior intervention plan, developing a written plan is not sufficient. The plan must be implemented.

Arizona guidance notes the same distinction and clarifies that a BCBA is not specifically required to conduct an FBA — the necessary expertise depends on the circumstances of the case.

Where ABA fits

Applied Behavior Analysis is the science and professional discipline concerned with relationships between behavior and environmental variables, and the systematic application of behavioral principles to socially significant outcomes. FBA, reinforcement, antecedent intervention, functional communication training, self-management, and prompting all draw on it.

But the terms are not interchangeable, and conflating them causes real confusion in IEP meetings. An FBA is not ABA therapy. PBIS is not ABA therapy. A school behavior plan is not automatically an ABA program. A family who has been told their child “needs ABA” and a team writing a function-based BIP are frequently talking about different services, in different settings, delivered by different people.

Contemporary behavior-analytic ethics emphasize stakeholder involvement, understandable communication, informed consent and assent where applicable, collaboration, cultural responsiveness, socially significant goals, and dignity. That framing matters for how a school states its own purpose:

The goal of function-based support is not compliance for its own sake. It is to understand what is getting in the way, teach skills the student can actually use, increase access and participation, and improve outcomes that are meaningful to the student, the family, and the school.

Using FarPoint’s tools

FarPoint provides free browser-based tools for the parts of this workflow that are structured enough to benefit from one. The Competing Behavior Pathway builder lays out the diagram above with your own case information and prints it for a team meeting. The Positive Behavior Support Plan tool works through the plan components in order.

Nothing is stored and no student data leaves the browser. These tools organize clinical reasoning and produce a document. They do not conduct the assessment, they do not assign function, and they do not substitute for direct observation, team process, or professional judgment about a specific student in a specific setting.

There is a through-line between this page and the CBM reference, and it is the point of both. In academic intervention: did we implement it, and did performance change? In behavior intervention: did we implement it, did the target behavior change, and did the replacement behavior increase?

Stop documenting that something was done. Start documenting whether it worked.

References

Primary sources for the requirements, frameworks, and findings described above.

Federal guidance

U.S. Department of Education, Office of Special Education and Rehabilitative Services, & Office of Elementary and Secondary Education. (2024, November). Using functional behavioral assessments to create supportive learning environments. Federal guidance on function-based assessment and support: observable behavior definitions, direct and indirect data collection, environmental variables, skill development, BSP/BIP implementation, and the use of FBA for students with and without disabilities whose behavior interferes with learning.

Implementation guidance

Center on PBIS. (2022). Tier 3 comprehensive functional behavior assessment (FBA) guide. University of Oregon. School-based guidance for conducting comprehensive FBAs, developing function-linked behavior support plans, and using data to evaluate intensive Tier 3 supports.
Center on PBIS. (2025, February). Tiered Fidelity Inventory 3.0 (TFI) manual. University of Oregon. Current PBIS systems-level fidelity measure spanning universal, targeted, and intensive implementation. It measures schoolwide implementation, not the fidelity of an individual student’s behavior plan.

FBA and BIP quality research

Iovannone, R., Moore, T. D., Williams, J. M., Sanchez, S., & Kauk, N. (2024). Are we on course yet? Functional behavior assessment and behavior intervention plan technical adequacy in schools. Behavioral Sciences, 14(6), 466. https://doi.org/10.3390/bs14060466 Evaluation of 135 FBAs and 129 BIPs from 13 school districts. Overall technical adequacy averaged roughly 40–50% of possible components, with specific weaknesses in operational definitions, hypothesis-linked intervention, progress monitoring, and implementation-fidelity planning.
Iovannone, R., Moore, T. D., Williams, J. M., Sanchez, S., & Kauk, N. (2024). Technical Adequacy Tool for Evaluation (TATE) [Unpublished measure]. University of South Florida. Released publicly by the Center on PBIS in January 2026 for evaluating completed FBAs and BIPs and identifying district-level improvement needs.

Educator training

The IRIS Center. (n.d.). Functional behavioral assessment (secondary): Identifying the reasons for student behavior. Vanderbilt University. Form versus function, reducing subjectivity, defining behavior, indirect and descriptive assessment, and hypothesis statements.
The IRIS Center. (n.d.). Behavioral intervention plans (secondary): Developing a plan to address student behavior. Vanderbilt University. Replacement behavior, antecedent intervention, reinforcement, progress monitoring, and implementation fidelity — including the instruction to address fidelity problems before concluding a plan needs changing.

State guidance

Arizona Department of Education. (n.d.). FAQs: Functional behavioral assessment and behavioral intervention plans. Addresses several common misconceptions directly: IDEA does not dictate an FBA format, Arizona does not require a BCBA to conduct every FBA, an FBA is not automatically required whenever behavior impedes learning, and the explicit requirements arise in particular disciplinary circumstances.
Arizona Department of Education. (2026, March 16). Behavior support. State resource hub linking PBIS, Basic FBA-to-BSP training, IRIS materials, federal behavior guidance, and IDEA discipline requirements.

Indirect assessment psychometrics

Paclawskyj, T. R., Matson, J. L., Rush, K. S., Smalls, Y., & Vollmer, T. R. (2000). Questions About Behavioral Function (QABF): A behavioral checklist for functional assessment of aberrant behavior. Research in Developmental Disabilities, 21(3), 223–229. https://doi.org/10.1016/S0891-4222(00)00036-6 Development study reporting test-retest, inter-rater, and internal consistency evidence.
Matson, J. L., Bamburg, J. W., Cherry, K. E., & Paclawskyj, T. R. (1999). A validity study on the Questions About Behavioral Function (QABF) scale: Predicting treatment success for self-injury, aggression, and stereotypies. Research in Developmental Disabilities, 20(2), 163–175. https://doi.org/10.1016/S0891-4222(98)00039-0 Validity study with 398 participants with intellectual disability. QABF findings contribute to a functional hypothesis and should be integrated with direct observation and other assessment data.

Experimental behavior analysis

Melanson, I. J., & Fahmie, T. A. (2023). Functional analysis of problem behavior: A 40-year review. Journal of Applied Behavior Analysis, 56(2), 262–281. https://doi.org/10.1002/jaba.983 Review of 1,333 functional analysis outcomes across 326 studies, covering evolving methods, multiple-function outcomes, and adaptations intended to improve efficiency and safety.
Heath, H., Jr., & Smith, R. G. (2019). Precursor behavior and functional analysis: A brief review. Journal of Applied Behavior Analysis, 52(3), 804–810. https://doi.org/10.1002/jaba.571 Methods for identifying lower-severity behaviors that reliably precede severe problem behavior, and precursor analysis as a potentially safer assessment approach.

Autism intervention synthesis

Hume, K., Steinbrenner, J. R., Odom, S. L., Morin, K. L., Nowell, S. W., Tomaszewski, B., Szendrey, S., McIntyre, N. S., Yücesoy-Özkan, S., & Savage, M. N. (2021). Evidence-based practices for children, youth, and young adults with autism: Third generation review. Journal of Autism and Developmental Disorders, 51, 4013–4032. https://doi.org/10.1007/s10803-020-04844-2 Synthesis of 972 articles identifying 28 focused intervention practices. It evaluates behavioral, developmental, and educational interventions — not a single comprehensive treatment model.

Professional ethics

Behavior Analyst Certification Board. (2020). Ethics code for behavior analysts (updated August 2024). Professional standards addressing scientifically grounded assessment and intervention, minimizing risk, prioritizing positive reinforcement, client and stakeholder involvement, informed consent and assent when applicable, and practicing within scope of competence.