Why Behavior Happens: Functions and Functional Assessment

The Central Idea

A behavior's function is the effect that keeps the behavior happening. Put simply, function is what the behavior gets the person, or gets them out of, in a way that makes it more likely to happen again.

Function is not what the behavior looks like. It is also not a guess about what the person "meant." Two people can both throw a pencil for completely different reasons.

To work toward the function, watch what happens just before and just after the behavior, many times. Repeated patterns help your team form a hypothesis, which is a best guess about the likely function. Patterns alone do not prove what keeps the behavior going. When it is needed, a functional analysis, led by someone trained in that method, can test the hypothesis.

Your supervisor uses function to choose a plan. Your accurate observations help them find it.

By the end of this lesson, you should be able to:

• Name the four common functions and match a new example to the most likely one.
• Explain why one form can serve different functions, and different forms can serve one function.
• Tell apart indirect assessment, direct descriptive (ABC) observation, and functional analysis.
• Record what you see instead of what you guess.
• Describe what an RBT does, and does not do, during a functional assessment.

Before you start: a behavior is something a person does that can be seen or measured. An RBT works under close, ongoing supervision from a supervisor who meets BACB requirements (a BACB-required supervisor).

The Four Common Functions

A consequence is what happens right after a behavior. If a consequence keeps the behavior going, it is reinforcing the behavior.

• Attention (social): Another person responds. They talk, look, scold, or comfort. A reprimand can still be attention.
• Access to items or activities (tangible): The behavior produces something the person wants, such as a toy, a snack, or a turn on a game.
• Escape or avoidance: The behavior ends something unpleasant (escape) or stops it from starting (avoidance). Examples include a hard task, a loud room, or a request.
• Automatic reinforcement: The behavior produces its own effect, and no other person needs to be involved. For example, it may produce a sound or a sensation the person likes, or reduce one they dislike.

Four contrasts: change one fact, change the function

1. Mo yells, and an adult walks over and chats. That suggests attention. Now suppose the adult silently hands Mo a puzzle and walks away. That suggests access.
2. Ari tears up a worksheet, and the worksheet is taken away. That suggests escape. Now suppose Ari tears it up as the worksheet is being brought over, and it never arrives. That suggests avoidance.
3. Jun hums loudly in a group, and peers laugh. That suggests attention. Now suppose Jun hums the same way while alone, with no one responding. That points toward automatic reinforcement. It does not prove it: no response from others is a clue, not a conclusion.
4. Lee grabs a ball from a peer and keeps playing with it. That suggests access. Now suppose Lee grabs the ball, an adult immediately takes it back and gives a long lecture, and grabbing keeps happening. Now attention is a serious possibility.

Same form, different functions: Humming served attention in one setting and possibly automatic reinforcement in another.

Different forms, same function: Pushing work away, saying "no," and putting one's head down can all end a task. If they do, they may all serve escape.

Do not assume a function from someone's diagnosis, age, or what "people like that usually do." Look at the person's own data instead.

Check 1

Priya bangs a cup on the table. Each time, a staff member refills it with juice without saying anything. Which function is most likely, and what single fact would make attention more likely?

Answer: Check 1

Access is most likely, because the behavior produces juice. Attention would become more likely if the staff member talked to Priya each time but did not refill the cup.

The tempting mistake is to say "attention" because a person was involved. Ask what changed after the behavior, not who was nearby.

How a Functional Assessment Gathers Evidence

A functional behavior assessment (FBA) collects information to identify a behavior's likely function. It uses up to three kinds of evidence:

• Indirect: Information that does not come from watching the behavior now. Interviews and rating questionnaires ask people who know the client to report what they have seen, so they depend on memory. Record reviews look at existing documents, such as incident reports or past notes. All of these are useful, but none is live observation of the behavior.
• Direct descriptive: Someone observes the behavior as it naturally happens and records it, often on an ABC chart.
• Functional analysis (FA): A planned experiment. Different conditions are deliberately arranged, and behavior is measured to test a possible function. This is used only when indicated. As the University of Minnesota Institute on Community Integration's guide Creating a Functional Analysis puts it, "A functional analysis is led by someone trained to use this experimental method of assessment." Because these conditions can bring out challenging behavior, they require trained leadership.

Reading an ABC chart

ABC stands for Antecedent (what happened just before), Behavior (what the person did), and Consequence (what happened just after).

What this suggests:

• Pushing the worksheet was followed twice by the work being removed. That points toward escape.
• Desk slapping was followed by the tech returning. That fits attention.
• Tapping happened with no one responding. That leaves automatic reinforcement as a possibility.

These are hypotheses: best guesses to test, not conclusions. The chart shows what happened together in time, not what caused what. A single row never settles a function.

Notice also that the antecedent is not the function. The worksheet appearing set the occasion for pushing it away. The removal of the worksheet is the consequence that may keep pushing going.

Record observations, not interpretations

• Write "Kai pushed the worksheet off the desk," not "Kai was frustrated." Frustration is an interpretation, not something you saw.
• A blank where you didn't observe is missing data, not zero. "Not observed" does not mean "did not happen."
• Counts need their time frame. Four events in 20 minutes is 4 ÷ 20 = 0.2 per minute. Four events in 60 minutes is 4 ÷ 60 ≈ 0.07 per minute. "4 times" means different things in each case, so always record the observation length.

The RBT's Role

You participate in assessment under your BACB-required supervisor. You take on an assessment task only after your supervisor confirms you have demonstrated competence in it. You:

• Observe and record accurately, using the assigned method.
• Carry out only the assessment steps you have been trained and assigned to do.
• Report concerns and unusual data to your supervisor promptly.

You do not:

• Design or start a functional analysis on your own.
• Hold back a safety response to "see what happens."
• Change a treatment plan because of a hunch, even a well-supported one.

Practice on Your Own

1. Rowan says "I'm done" and walks away whenever the vacuum turns on in the next room. The vacuum keeps running. Rowan stays in a quiet room. What function is most likely? Is this escape or avoidance?

2. A caregiver completes a 16-question rating form about a client's behavior. What kind of assessment evidence is this?

3. During an ABC observation, you suspect attention. You consider ignoring the client for ten minutes to test your idea. What should you do instead?

4. A coworker's note reads: "Client was bored. No aggression." The coworker observed for the first 15 minutes of a 2-hour session. Name two problems.

5. Two clients each scream. For one, screaming is followed by a peer handing over a tablet. For the other, screaming continues when they are alone and no one responds. Explain why the same behavior is listed under different functions.

6. While a caregiver is on the phone, Sam taps the caregiver's arm, calls "Hey!" loudly, and tosses a sock at the caregiver's feet. Each time, the caregiver pauses the call and talks to Sam. These three behaviors look very different. Could they share one likely function? Explain what you would look at.

Answer: Practice on Your Own

1. The most likely function is escape. The noise was already present, and leaving ended Rowan's contact with it. It would be avoidance if Rowan left before the vacuum started. This is still a hypothesis that needs more observations.

2. This is indirect assessment. The caregiver is reporting past behavior from memory, not observing it live or running an experiment.

3. Keep recording and tell your supervisor what you noticed. Deliberately arranging conditions to test a function is functional analysis, which a trained person designs and leads. Your idea is useful information, not permission to experiment.

4. "Bored" is an interpretation, not an observation. "No aggression" covers only 15 minutes, so the other 105 minutes are missing data, not zero.

5. Function depends on what keeps the behavior going, not on what it looks like. For the first client, the effect is access to the tablet. For the second, the behavior continues with no social response, which suggests automatic reinforcement.

6. Yes. Look at the consequence, not the form. Tapping, calling out, and tossing a sock are all followed by the same thing: the caregiver stops and talks to Sam. That shared consequence suggests all three may serve attention. The tempting mistake is to treat them as three separate problems because they look different. This is a hypothesis; more observations would show whether the pattern holds.

Recap

• Function is the effect that maintains a behavior. It is not the behavior's appearance or a guess about intent.
• The four common functions are attention, access, escape/avoidance, and automatic.
• One form can serve different functions, and different forms can share one function. Compare consequences, not appearances.
• Indirect assessment means reports from people or reviews of existing records. Direct descriptive assessment means live ABC recording. Functional analysis is a supervised experiment led by someone trained in the method.
• ABC patterns suggest hypotheses, not causes. One event never decides a function.
• Record what you see, mark missing time as missing, and include the time frame.
• RBTs observe, record, and carry out assigned steps after their supervisor confirms competence. They do not design assessments, skip safety responses, or change plans.

Sources: Sam & AFIRM Team (2024), Functional Behavior Assessment, UNC Frank Porter Graham Child Development Institute. University of Minnesota Institute on Community Integration, Creating a Functional Analysis. BACB, RBT Test Content Outline (3rd ed.). BACB, RBT Ethics Code (2.0).

Independent study resource

This study resource is not affiliated with, sponsored by, or endorsed by the Behavior Analyst Certification Board (BACB). BACB, Registered Behavior Technician and RBT are registered trademarks of the Behavior Analyst Certification Board. Source citations do not imply endorsement by their authors or publishers.