B.3 Functional assessment

What the outline says

Task B.3 in content area B, Behavior Assessment (8 of the 75 scored questions): "Participate in components of functional assessment procedures (e.g., descriptive assessment, functional analysis)."

Why function matters

Behavior that keeps happening is being reinforced. The common functions are social attention, escape from or avoidance of something unpleasant (such as demands), access to tangible items or activities, and automatic reinforcement, in which the behavior produces its own reinforcement, such as a sensation.

Function is not the same as form. Two clients may both bite their hands, one to escape hard tasks and the other because the behavior is automatically reinforced. One client may scream, drop to the floor and throw a pencil, and all three behaviors may end a worksheet. Because a plan has to fit the function, for example by teaching a replacement behavior that gets the same reinforcer, the function has to be identified first. A functional behavior assessment (FBA) is the overall process for doing this.

The three types of methods

• Indirect assessment: interviews, questionnaires and rating scales completed by people who know the client. These are quick, but they rely on memory and impressions, so they are usually followed by direct methods.

• Descriptive assessment: direct observation of the behavior in the client's everyday settings, without changing anything. Examples are ABC recording and scatterplots. It shows which events tend to occur with the behavior, which is a correlation, not proof of cause.

• Functional analysis (FA): the team arranges antecedents and consequences on purpose in test and control conditions and measures the behavior in each. Because the conditions are controlled, an FA can demonstrate which consequence maintains the behavior.

Collecting ABC data

ABC stands for antecedent (what happened just before), behavior (what the client did) and consequence (what happened just after). In narrative recording you write a short description of each; in structured recording you check preset categories the supervisor has listed. Note the time, activity and people present as well.

Steps for good ABC data:

• Know the operational definition and record only what meets it. If elopement is defined as moving more than 3 meters from the area, a 1-meter walk is not recorded.

• Record each occurrence as soon as possible, not from memory at the end of the day.

• Describe events, not guesses. Write "adult walked over and said, 'What's wrong?'" rather than "adult reinforced the screaming," and "hit the table with an open hand 3 times" rather than "got angry." The absence of something is also an antecedent: "alone with puzzle; no adult interaction for about 5 minutes" is far more useful than "nothing happened."

• Do not step in. In a descriptive assessment the aim is to see what normally happens, including how caregivers respond.

• Record across routines and days so the sample represents the client's typical day, and note if the client seems to react to being watched (reactivity).

Remember the limits. Adults often respond to problem behavior whatever its function, so a reprimand that follows nearly every outburst does not prove that attention maintains it. A scatterplot can show that behavior clusters at 11:30 every day, which tells the team where to look next, but not why it happens.

Functional analysis conditions

A typical FA compares several test conditions with a control condition. You may be asked to run them, so know what each one looks like:

• Attention: the adult is busy (for example, reading) and gives brief attention, such as a reprimand or statement of concern, only after the target behavior. Going without attention makes attention more valuable, so the condition tests whether attention maintains the behavior.

• Escape (demand): the adult presents tasks with prompts and removes the task briefly after the target behavior. If behavior increases because it removes tasks, that is negative reinforcement.

• Tangible: a preferred item is removed and returned briefly after the target behavior.

• Alone or ignore: no social consequences are provided. In an ignore condition, an adult stays in the room for safety but does not respond. Behavior that persists here suggests automatic reinforcement.

• Play (control): preferred items and attention are freely available and no demands are placed. Problem behavior is expected to be low, and the other conditions are compared with it.

Reading results: behavior that is high only in the demand condition suggests escape. Behavior that is high in every condition, including alone, suggests automatic reinforcement. The supervisor makes these interpretations; you should understand them.

Running conditions with fidelity and safety

Each condition tests one consequence, so procedural fidelity is essential. Comforting a client while removing a task adds attention to the escape condition, and answering an appropriate question in the attention condition gives attention for a behavior other than the target. Both blur the results. Deliver exactly the consequences the protocol lists and nothing else.

Because FA conditions are designed to evoke problem behavior, the supervisor sets safety rules, such as when to end a session. Follow them every time; a complete data set never matters more than the client's safety. Document and report anything unplanned, such as another staff member walking in and talking to the client.

If sessions differ in length (for example, one was ended early under a safety rule), compare rates, not raw counts. For example, 12 responses in a 10-minute attention session is 1.2 per minute, and 9 responses in a 5-minute demand session is 1.8 per minute, so the demand session had the higher rate. A second observer may record at the same time to check interobserver agreement. If two observers' records differ a lot, tell the supervisor rather than averaging them or choosing one.

Reporting and boundaries

Report variables that can change behavior before or during an assessment: illness, a new medication, poor sleep, schedule changes. Treat ABC sheets and session videos as confidential client records. If a teacher or caregiver asks why the behavior happens, explain that the supervisor will review the data and share the findings. Choosing conditions, interpreting the graph and writing the behavior plan are the supervisor's responsibilities.

Common mistakes

• Confusing correlation with cause. ABC data and scatterplots show patterns; only a functional analysis tests what maintains behavior.
• Writing interpretations on the ABC sheet ("wanted attention," "was bored"). Record observable events and conditions.
• Adding consequences in FA conditions, even kind ones. Extra attention or items change what the condition tests.
• Ignoring safety rules to finish a session. End the session when the rules say to, and report it.
• Telling teachers or caregivers what the function is. Pass the question to the supervisor.

How it's tested

Expect to identify antecedents, behaviors and consequences in short scenarios, to choose objective ABC entries, to match each method (indirect, descriptive, functional analysis) to its description, and to name FA conditions from a description of what the adult does. Some items ask what results suggest about function or why an FA is stronger evidence than ABC data. Scenario items test procedural fidelity, safety and reporting. Choose the option in which the RBT follows the protocol exactly and leaves design and interpretation to the supervisor.