When Functional Assessment Clues Don’t Match: How Teams Build a Careful Hypothesis

A caregiver says a behavior happens when attention is unavailable. A teacher sees it most often during difficult work. Direct observation shows both patterns, but neither one happens every time.

Which explanation is correct?

Sometimes the honest answer is: **the team does not know yet.**

Conflicting information does not mean the assessment has failed. It may mean the behavior has more than one function, the settings are different, an important condition has not been observed, or the team needs a clearer definition and more focused data.

A functional assessment should reduce guessing, not hide uncertainty. This guide explains how teams can compare mixed clues, write a careful hypothesis, and decide what information is still needed. It is general education, not individualized assessment, treatment advice, or legal guidance. Assessment and any experimental analysis of dangerous behavior should be designed and overseen by qualified professionals.

A Hypothesis Is an Informed Explanation, Not a Verdict

In functional assessment, a hypothesis describes the conditions in which a behavior is more likely and the outcome that may be maintaining it.

A useful hypothesis might say:

> When independent writing is presented without a model or help, pushing materials away and leaving the seat are more likely. These responses are often followed by a pause in the task or adult assistance. The behavior may help the learner delay difficult work and obtain support.

Notice the cautious language: **more likely**, **often**, and **may**. The statement does not claim to know the learner's intent. It connects observable events and leaves room for the hypothesis to change.

For a simple introduction to common behavioral functions, see What Is the Function of Behavior?. For the difference between assessment and the support plan that follows, see FBA vs BIP: A Parent-Friendly Guide.

Where Functional Assessment Clues Come From

Teams often gather several kinds of information.

Interviews and rating tools

Caregivers, educators, direct staff, and the learner can describe routines, patterns, preferences, communication, health changes, and situations an observer may not see. This indirect information is especially useful for identifying questions and choosing times or settings to observe.

An interview reflects the person's experience. It is valuable, but it is not the same as directly observing the behavior. Two people may also see different patterns because they spend time with the learner in different settings.

Records and existing data

Attendance, incident reports, work samples, previous plans, health information shared through appropriate channels, and earlier data may show changes over time. Records can add context, but their usefulness depends on how clearly and consistently information was recorded.

Direct observation

The assessor watches what happens before and after the behavior in everyday routines. ABC data is one way to organize these observations.

Direct observation shows what happened during the observation period. It does not automatically prove that the event before the behavior caused it or that the event afterward is maintaining it.

Functional analysis

A functional analysis systematically changes selected environmental conditions while behavior is directly measured. Its purpose is to test a specific question rather than rely only on correlation.

A functional analysis is not a do-it-yourself activity for families, teachers, or untrained staff. The need, design, consent process, safety protections, and interpretation require professional judgment. Not every situation calls for the same assessment method.

Why Clues May Point in Different Directions

Conflicting results can happen for understandable reasons.

People are describing different behaviors

One person may count yelling, crying, and leaving as one response. Another may record only leaving the area. If the team is not discussing the same observable behavior, the patterns will not line up.

The settings really are different

Behavior can serve one function during academic work and another during free time. It can also change depending on the person present, the difficulty of the task, communication options, noise, predictability, or how adults usually respond.

The learner has more than one way to meet the same need

A learner may ask for help in one setting and push materials away in another. If the effective communication response is available only sometimes, the behavior pattern may look inconsistent.

The behavior may have more than one function

The same form of behavior can produce different outcomes across situations. Leaving a room may help someone escape noise during an assembly and reach a preferred adult during another routine.

The relevant situation was not observed

Low-frequency behavior or behavior tied to a very specific event may not occur during a short observation. Adults may also change a routine when an observer is present.

The data show correlation, not function

Attention may follow a behavior because a caring adult responds to distress. That pattern alone does not show that attention is the reason the behavior continues. A difficult task may happen before behavior without being the only relevant factor.

Important setting events may be missing

Sleep disruption, illness, pain, hunger, medication changes, stressful events, or changes in routine can alter how a learner responds. ABA professionals should document relevant patterns and refer rather than diagnose. See Medical Causes That Can Look Like "Problem Behavior" for more on this boundary.

Expectations and bias can shape interpretation

Words such as "attention-seeking," "defiant," or "unmotivated" can make an assumption sound like a fact. Cultural expectations, adult stress, and prior reports may influence what people notice. Observable language helps the team return to what happened.

A Worked Example: Three Sources, Three Stories

Imagine a student who closes a laptop, puts their head down, or leaves the desk during class.

• A caregiver reports that the student does similar things when adults are busy, suggesting attention.
• A teacher reports that it usually happens during writing, suggesting escape from difficult work.
• Several observations show the behavior during independent writing. The teacher then approaches, talks with the student, and reduces or helps with the task.

The clues do not require the team to choose attention or escape immediately. Adult interaction and task relief occur together. The team might ask:

• Does the behavior happen during easy independent work?
• Does it happen during difficult work when help is already available?
• Does it happen when adult attention is limited but no task is present?
• Can the student ask for help, a model, a pause, or adult connection efficiently?
• Are writing, typing, language, motor, or other skill demands contributing?
• What happens when the task is broken into a smaller first step while attention stays similar?

The next step is not to vote on the function. It is to collect information that separates the competing explanations.

How Teams Can Work Through Conflicting Information

1. Confirm the behavior definition

Write what a person can see or hear. Include examples, nonexamples, when the behavior begins and ends, and whether different responses should be measured separately.

"Shuts the laptop screen more than halfway or moves more than six feet from the assigned desk within 30 seconds of a work direction" is clearer than "avoids work."

2. Compare the conditions, not just the conclusions

Instead of listing each person's guessed function, organize the raw information:

• What happened before the behavior?
• What response occurred?
• What happened immediately afterward?
• What task, person, setting, or communication option was present?
• When was the behavior absent?
• Did the learner have an effective alternative response?

This can reveal that two reports are not truly contradictory. They may describe different routines.

3. Check the quality and coverage of each source

Ask how many examples support the pattern, when the information was collected, and whether the relevant situation occurred. A detailed interview can identify important possibilities. Repeated direct observations can show consistent associations. Neither should be treated as infallible.

More data is not always better data. Ten vague notes may be less useful than four clear observations across the situations the team needs to compare.

4. Look for both occurrence and nonoccurrence

The team learns from times when the behavior does not happen.

If the learner completes writing when choices, a model, or help are available, that difference may be more informative than another record of behavior during unsupported writing. If the behavior occurs when attention is unavailable during play but not when attention is unavailable during a preferred independent activity, the context matters.

5. Write competing hypotheses

When two explanations remain plausible, document both.

• **Hypothesis A:** The behavior may produce a delay or reduction in difficult work.
• **Hypothesis B:** The behavior may produce adult help or interaction.
• **Combined possibility:** The learner may be escaping unsupported work specifically by obtaining adult assistance.

Writing alternatives protects the team from settling too quickly on the first familiar explanation.

6. Collect targeted information

Plan observations that are likely to clarify the question. This might mean observing easy and difficult tasks, supported and independent work, or routines with different levels of adult availability.

This does not mean casually provoking dangerous behavior. Qualified professionals should determine whether additional descriptive observation, skill assessment, consultation, or a carefully designed functional analysis is appropriate.

7. Include the learner's perspective

When possible, ask the learner what feels difficult, uncomfortable, confusing, or helpful. Use the person's preferred communication method. Learner report may not answer every functional question, but excluding it can leave out important information about pain, effort, fear, sensory experience, social context, or what support feels acceptable.

8. Match support to the current evidence and monitor it

While assessment continues, teams can often use low-risk, dignity-centered supports such as clarifying expectations, making communication easier, adjusting unreasonable demands, offering choices, improving predictability, and reinforcing useful replacement skills.

The team should state what it expects the support to change and review the data. A helpful response to support adds information, but it does not prove the hypothesis by itself.

What a Careful Hypothesis Statement Includes

A useful statement usually names:

• the specific behavior
• the situations in which it is more likely
• relevant events before the behavior
• the outcomes that regularly follow
• the possible function or functions
• the evidence supporting the statement
• important uncertainty or missing information

For example:

> Across four classroom observations, leaving the work area was most likely during independent tasks with multiple written responses. It was usually followed by adult help and a pause in work. Interviews also identified writing as difficult, while attention-only situations were not observed. Current information suggests the behavior may help the student access assistance and delay difficult writing. Observation during nonwriting tasks and attention-only situations is still needed.

That statement gives the team something testable. It is stronger than "The behavior is escape-maintained" because it shows the evidence and the limits.

What Families, Educators, and Direct Staff Can Contribute

People who know the learner well are essential partners. They can:

• describe specific examples rather than only choosing a function
• identify routines where the behavior does and does not happen
• share changes in communication, sleep, health, medication, stress, or daily routines through appropriate private channels
• explain what adults usually do next
• point out skills or supports that make the routine easier
• report whether a proposed replacement response is realistic
• note when an assessment summary does not match their experience
• ask how the hypothesis will guide support and how the team will know whether it fits

School-based procedures and legal requirements vary. Families can use ABA and IEPs: How Behavior Goals Fit With School Support as a starting point for broader school-team questions, while seeking local guidance for their situation.

When the Team Should Slow Down or Broaden the Assessment

More behavioral data is not always the right next step. Pause and consider additional expertise when:

• behavior changes suddenly or may involve pain, illness, medication, sleep, feeding, or another health concern
• there is significant risk of injury
• trauma, bullying, discrimination, grief, or mental-health concerns may be involved
• communication, academic, motor, sensory, or environmental barriers have not been assessed
• data collectors do not agree on what counts as the behavior
• the behavior occurs only in one person's presence or one setting
• proposed assessment conditions would create unacceptable risk or distress
• the team cannot explain how the learner's dignity, assent, privacy, and safety will be protected

A careful assessment recognizes professional boundaries. Collaboration may be more useful than forcing every concern into a single behavioral explanation.

Questions to Ask When the Clues Do Not Match

Families and team members can ask:

• Are we all measuring the same behavior?
• Which observations support each hypothesis?
• When does the behavior not occur?
• Did each source have an opportunity to see the relevant routine?
• Could two functions be operating in different settings?
• Are adult help and task removal happening at the same time?
• What communication or coping response is available to the learner?
• Have skill difficulty, pain, health, sensory needs, stress, and environmental barriers been considered?
• What information would help us distinguish the competing explanations?
• Is a functional analysis appropriate, and who is qualified to design it?
• How will we check whether the support plan matches the assessment?
• When will the team review and revise the hypothesis?

Final Thought

Functional assessment is a process of asking better questions, not attaching a permanent label to a learner.

When clues conflict, a thoughtful team does not force certainty. It defines the behavior clearly, compares contexts, listens to the learner and the people who know them, tests competing explanations carefully, and documents what remains unknown.

A hypothesis should become more useful as evidence improves. When it does not fit, revising it is not failure. It is responsible assessment.