If you work in ABA or special education, you have probably heard some version of this:
“We need to figure out why the behavior is happening.”
That is the job of a functional behavior assessment.
An FBA is a systematic way of looking at behavior in context so we can understand what is setting it off, what happens afterward, and what the person may be getting or avoiding by engaging in that behavior.
The goal is not to assign a label like “attention-seeking” or “sensory.”
The goal is to build a hypothesis that is specific enough to guide intervention.
A good FBA should eventually let the team say something like:
When this situation happens, this person does this behavior, and the behavior works because it produces this outcome.
That one sentence becomes the foundation for the behavior intervention plan.
And if that sentence is wrong, everything built on top of it is at risk of being wrong too.
What Does an FBA Actually Answer?
At its core, an FBA asks one question:
What is this behavior doing for the person?
Behavior that continues over time is usually behavior that works in some way.
It may help the person:
- Get attention
- Escape or avoid something difficult or unpleasant
- Access an item or activity
- Produce or reduce some kind of internal or sensory stimulation
Those are the four functions most clinicians learn first.
Researchers often organize the same idea more broadly as obtaining versus escaping, crossed with social versus non-social reinforcement.
The exact label matters less than understanding the contingency.
What happens before the behavior?
What does the behavior look like?
What happens immediately after?
And why would that consequence make the behavior more likely to happen again?
That is the real work.
Behavior Can Have More Than One Function
This is an important point because teams sometimes force behavior into one box too quickly.
Across 981 published functional analyses, escape was the most common identified function, accounting for about 32% of differentiated outcomes. Attention accounted for about 22%, tangible access about 11%, automatic reinforcement about 16%, and nearly 19% of behaviors were maintained by more than one contingency.
That means two things.
First, escape is more common than many adults realize.
Behavior that looks “attention-seeking” may actually be helping a student delay or avoid a task.
Second, roughly one in five behaviors in that literature had multiple functions.
So if the data point to more than one maintaining variable, the plan has to address more than one.
What Should the Final FBA Hypothesis Look Like?
The end product should be clear enough that the team can act on it.
A strong hypothesis usually includes:
- The routine or setting
- The immediate antecedent
- The behavior
- The consequence
- The likely function
A practical template is:
During [routine], when [antecedent], the person engages in [behavior], and this typically results in [consequence]. Therefore, the behavior is likely maintained by [function].
That is much more useful than writing:
“Behavior is escape-maintained.”
Because the context matters.
Escape from what?
Under which conditions?
What does the behavior actually look like?
What consequence follows it?
The more specific the hypothesis, the easier it becomes to design a useful intervention.
When Is an FBA Required in Schools?
In schools, there are both legal triggers and clinical reasons to conduct an FBA.
Under IDEA, one of the clearest triggers is discipline.
If a student with a disability is removed for more than 10 consecutive school days, or a pattern of removals becomes a change of placement, the team must conduct a manifestation determination.
If the behavior is found to be a manifestation of the student’s disability, the IEP team must conduct an FBA and implement a BIP if one has not already been completed, or review and revise the existing BIP.
Outside of discipline, IDEA also requires IEP teams to consider positive behavioral interventions and supports when behavior interferes with learning.
An FBA is often the best way to determine what those supports should actually look like.
IDEA Does Not Define One Universal FBA
This is where people sometimes get confused.
Federal law does not provide one universal checklist that defines exactly what every FBA must include.
The Department of Education has instead described the common features of a sound FBA:
- A clear description of the behavior
- Data collection
- A review of function
- A connection to skill development and intervention
States may add their own requirements.
New York, for example, requires direct observation, multiple informants, record review, baseline data across settings and times, and attention to antecedents, consequences, skills, and preferences.
So if you are working in a school system, know the federal requirements.
But know your state requirements too.
FBAs Are Not Only for Students With IEPs
Another misconception is that an FBA is only a special education tool.
It is not.
The Department of Education has stated that FBAs can be used for any student whose behavior interferes with learning, including students supported through MTSS or PBIS frameworks.
That matters because we should not have to wait until behavior becomes severe before we try to understand it.
What About Clinical ABA?
In clinical ABA, the structure is different, but the expectation is similar.
The BACB Ethics Code requires behavior analysts to select and design assessments before designing interventions, and interventions must be based on assessment results.
Payers often expect the same thing.
Treatment plans commonly require functional assessment results, operational definitions, hypotheses about function, baselines, and measurable behavior-reduction goals.
So whether the setting is a school or clinic, the principle is the same:
We should not be trying to change behavior before we understand what is maintaining it.
There Are Three Levels of Functional Assessment
Not every FBA needs the same amount of depth.
I think of functional assessment as a continuum.
At one end, we ask people what they have noticed.
In the middle, we directly observe the behavior in the environment where it occurs.
At the most intensive end, we experimentally test the variables that may be maintaining it.
Each level gives us a different degree of confidence.
Level 1: Indirect Assessment
Indirect assessment means asking people who know the individual what they have observed.
That may include interviews or rating scales such as:
- Functional Assessment Interview
- FACTS
- QABF
- FAST
- Motivation Assessment Scale
These tools can be very useful for generating hypotheses.
But they do not confirm function.
That distinction matters.
If a teacher says:
“He does it for attention.”
That is useful information.
But it is still a hypothesis.
If a FAST score points to escape, that is useful information.
But it is still a hypothesis.
Indirect tools should help us decide what to observe next.
They should not usually be the entire FBA.
Why Indirect Tools Need Caution
The source makes an important point about psychometric limits.
In one FAST validation study, agreement between raters was about 72% item by item, agreement on the overall outcome was about 65%, and the FAST matched functional analysis results in about 64% of cases.
That is useful for hypothesis generation.
It is not strong enough to justify treating an indirect rating scale as proof of function.
So when I see an FBA that consists of a FAST and one interview, I do not think:
Assessment complete.
I think:
We have a starting hypothesis. Now we need to watch what actually happens.
Level 2: Direct Descriptive Assessment
This is where the assessment becomes much more informative.
Direct descriptive assessment means observing the behavior in the settings where it naturally occurs.
No variables are intentionally changed.
The goal is to see what tends to happen before and after the behavior.
Common methods include:
ABC Recording
ABC stands for:
Antecedent. Behavior. Consequence.
For each occurrence, record:
What happened immediately before?
What did the person do?
What happened immediately after?
Over time, patterns emerge.
Scatterplots
A scatterplot tracks when behavior occurs across periods of the day.
This can be surprisingly useful.
Sometimes the biggest clue is not what happens immediately before the behavior.
It is when and where the behavior clusters.
Maybe it happens almost entirely during independent work.
Maybe it spikes before lunch.
Maybe it is rare when one particular staff member is present.
Those patterns can point us toward routines and setting events that interviews missed.
Conditional Probability
This looks more closely at how often a consequence follows a behavior compared with how often that consequence occurs at other times.
If task removal follows the behavior 80% of the time but tasks are rarely removed otherwise, that tells us something important.
Direct observation is where we begin checking the interview hypothesis against reality.
Level 3: Functional Analysis
A functional analysis, or FA, goes one step further.
Instead of simply observing naturally occurring events, the clinician deliberately arranges conditions and measures the behavior.
The classic functional analysis model includes conditions designed to test variables such as:
- Attention
- Escape from demands
- Access to tangibles
- Automatic reinforcement
- A control or play condition
If behavior reliably increases in one condition compared with control, that helps identify the maintaining function.
This is much stronger evidence because we are experimentally testing the hypothesis.
But it is also more resource-intensive and requires training, consent, and safety planning.
Functional Analysis Is Not Necessary for Every FBA
I think this point is important because “functional assessment” and “functional analysis” are sometimes used as though they mean the same thing.
They do not.
A functional analysis is one possible method within the larger FBA process.
Many school and clinic cases can be answered well with interviews plus direct observation.
A functional analysis may be especially useful when:
- Descriptive data are unclear
- The behavior is dangerous enough that a wrong hypothesis has significant consequences
- A function-based plan has already failed
- Greater experimental certainty is clinically necessary
There are also practical variants, including brief, trial-based, latency-based, and interview-informed analyses.
The point is not to use the most complicated assessment possible.
The point is to use enough assessment to answer the question with appropriate confidence.
Brief FBA or Comprehensive FBA?
Not every behavior needs the same assessment intensity.
For a relatively low-risk behavior that occurs in one or two predictable routines, a brief FBA may be enough.
For example:
- Calling out
- Leaving a seat
- Work avoidance
- Minor disruption
A brief FBA may involve a focused staff interview and direct observation in the identified routine.
For chronic, high-intensity, dangerous, or cross-setting behavior, the assessment needs to be more comprehensive.
That may include:
- Records review
- Multiple interviews
- Family input
- Student input when appropriate
- Direct observation across settings
- Ongoing baseline data
- Functional analysis if needed
The bigger the risk of getting the function wrong, the stronger the assessment should be.
How I Think Through an FBA Step by Step
The actual process can be organized in many ways, but this is the sequence I find most useful.
1. Build the Right Team
You need people who know the individual well.
You need someone who understands behavior assessment.
And you need people who understand the environment where intervention will actually happen.
That may include:
- Family
- Teachers
- Aides
- BCBAs
- School psychologists
- Behavior specialists
- The student, when appropriate
Behavior does not happen in a vacuum.
The team should not either.
2. Define the Behavior Clearly
This is one of the most important steps.
“Noncompliant.”
“Disruptive.”
“Anxious.”
“Aggressive.”
Those words may describe our interpretation.
They do not necessarily tell us what to observe.
A useful operational definition tells us exactly what the person does.
It should be observable and measurable.
It should include examples and non-examples.
And someone unfamiliar with the student should be able to recognize the behavior from the definition alone.
3. Review What Is Already Known
Look at:
- Previous evaluations
- Incident reports
- Attendance
- Academic history
- Medical information when relevant
- Previous behavior plans
- Previous FBAs
Sometimes the most important clue is already sitting in the record.
4. Interview the People Who See the Behavior
Ask:
When does it happen?
When does it not happen?
What usually comes before it?
What happens afterward?
Who is usually present?
What has already been tried?
What does the person seem to gain or avoid?
These interviews help shape the first hypothesis.
5. Watch the Behavior Directly
This is where the hypothesis meets reality.
Observe the routines that interviews identified.
Collect ABC data.
Use a scatterplot if time-of-day patterns matter.
Establish a baseline using the dimension that actually matters:
- Frequency
- Duration
- Latency
- Intensity
Do not settle for:
“It happens a lot.”
We need numbers.
6. Conduct a Functional Analysis if Needed
If the picture is still unclear, or the stakes are high, experimentally test the hypothesis.
7. Write the Hypothesis
Now bring the pieces together.
What setting events make the behavior more likely?
What immediate antecedent tends to trigger it?
What does the behavior look like?
What consequence follows?
What function does the behavior serve?
If the data support more than one function, say so.
8. Translate the Findings Into Recommendations
This is where the FBA starts to become useful.
The report should identify:
- A functionally equivalent replacement behavior
- Antecedent changes
- Reinforcement strategies
- Response strategies
- Data to collect
- When the team should review the plan
The FBA should hand the intervention team a clear behavioral map.
What Belongs in the Written FBA Report?
A strong report should generally include:
- Identifying information
- Team members
- Reason for referral
- Relevant records review
- Operational definitions
- Indirect assessment results
- Direct observation data
- Functional analysis results if one was conducted
- Setting events
- Antecedents
- Consequences
- Baseline data
- Hypothesis statement
- Competing behavior pathway
- Preference information
- Recommendations
The exact format can vary.
The important part is that another clinician or team member should be able to read the report and understand how you reached the hypothesis.
Common FBA Mistakes
Some mistakes come up again and again.
Stopping at the Interview
An interview is useful.
A rating scale is useful.
Neither is enough by itself in most cases.
If you have not watched the behavior in the environment where it occurs, you are missing one of the most important parts of the assessment.
Having No Real Baseline
“Frequent” is not a baseline.
“Severe” is not a baseline.
“Usually happens during math” is not a baseline.
If frequency matters, count frequency.
If duration matters, measure duration.
The baseline needs to match the behavior definition and the question you are trying to answer.
Defaulting to “Sensory”
Automatic reinforcement is real.
But it is also sometimes used as the explanation when we have not found a clearer social function.
Before concluding that a behavior is automatically reinforced, make sure the data actually support that conclusion.
Ignoring When the Behavior Does Not Happen
Sometimes the most useful information in an FBA is absence.
If behavior happens during independent seatwork but almost never during small-group instruction, that contrast is telling you something important.
What is different?
Adult proximity?
Task difficulty?
Rate of reinforcement?
Predictability?
Opportunity for help?
Do not only study the moments when behavior occurs.
Study the moments when it does not.
Assuming One Behavior Has One Function
If the data suggest multiple functions, build that into the hypothesis.
A plan that treats only one maintaining variable can leave the other one completely intact.
Treating the FBA as Permanent
Behavior changes.
Environments change.
People change.
Skills change.
If a plan worked for six months and suddenly stops working, the answer is not always:
Use a bigger reinforcer.
Sometimes the variables maintaining the function have changed.
Sometimes the antecedents have changed.
Sometimes the old assessment no longer describes the current behavior.
That is when we return to the FBA.
An FBA Is Not the Same as Measuring Overall Development
This distinction matters in ABA.
An FBA answers a very specific question:
Why is this behavior happening?
That is valuable.
But it does not tell us everything about the learner’s development.
It does not tell us how adaptive functioning compares with peers.
It does not tell us how broad developmental skills are changing.
It does not replace skills assessments or other standardized measures.
A strong clinical record needs multiple kinds of data because each one answers a different question.
Frequently Asked Questions
What is a functional behavior assessment?
A functional behavior assessment is a systematic process for identifying why a behavior occurs.
It looks at the situations that precede behavior, what follows it, and the broader setting events that may make the behavior more likely.
The goal is to develop a function-based hypothesis that can guide intervention.
What are the four functions of behavior?
A common framework includes:
- Attention
- Escape or avoidance
- Access to tangibles or activities
- Automatic or sensory reinforcement
A behavior can also have more than one function.
Who can conduct an FBA?
There is no single federal credential required for every FBA.
The Department of Education expects the assessment to be conducted by professionals with the necessary skills, training, and knowledge, in collaboration with families.
In practice, BCBAs, school psychologists, and behavior specialists often lead the process.
Experimental functional analyses require specific behavior-analytic training.
How long does an FBA take?
There is no single answer.
A brief FBA focused on one routine may be completed relatively quickly.
A comprehensive FBA may take days or weeks because it includes multiple interviews, records review, and observations across settings.
If the FBA is part of a formal special education evaluation, the applicable evaluation timeline applies.
Is parent consent required?
It depends on the purpose of the FBA.
Federal guidance states that consent is required when the FBA is part of an initial evaluation or reevaluation, or when it functions as the evaluation.
Consent is not federally required for every instructional or screening FBA within MTSS or PBIS, although state requirements may be stricter.
What is the difference between an FBA and a functional analysis?
An FBA is the entire assessment process.
It may include:
- Interviews
- Rating scales
- Record review
- Direct observation
- Functional analysis
A functional analysis is one specific experimental method within that process.
Every FA can be part of an FBA.
Most FBAs do not require an FA.
What does an FBA hypothesis statement look like?
A good hypothesis describes the routine, antecedent, behavior, consequence, and likely function.
For example:
During independent seatwork, when a student is given a difficult multi-step worksheet with no adult nearby, the student leaves the assigned area, and the task is delayed or removed. The behavior is likely maintained by escape from difficult academic demands.
Can a student without an IEP receive an FBA?
Yes.
Federal guidance states that FBAs can be used to support any student whose behavior interferes with learning, including within MTSS and PBIS frameworks.
How often should an FBA be updated?
There is no universal interval.
I would revisit the FBA when:
- The behavior changes substantially
- A new behavior emerges
- The environment changes
- The team changes significantly
- A previously effective intervention stops working
- New data no longer fit the existing hypothesis
The FBA should remain a working clinical document, not something we complete once and never revisit.
The Takeaway
A functional behavior assessment should leave the team with more than a label.
It should leave them with an explanation they can use.
When does the behavior happen?
What tends to set it off?
What does the person gain or avoid?
What evidence supports that conclusion?
We get there by defining the behavior clearly, talking with the people who know the person, observing the behavior directly, collecting a real baseline, and testing the hypothesis when the situation warrants greater certainty.
Then we use that information to build an intervention that actually fits the function.
Because the purpose of an FBA is not to produce another report.
It is to understand behavior well enough that we can respond more effectively.
