Skip to content

Functional Behavior Assessment

Principle · All principles
Evidence · 16 claims (11 for, 5 mixed) · 23 studies (8 quant-synthesis, 7 causal, 7 review, 1 theoretical), q1–q4 · 4 of 23 report an effect size · 11 claims rest on one study

Conditional relationship

This page owns finding out what sets off and keeps up one learner's recurring problem behavior, and choosing the response from that. For a learner whose disruptive or interfering behavior keeps recurring although ordinary classroom management is in place, recording when it happens, what comes just before it (the trigger) and what follows it (what the behavior gets the learner, such as attention or escape from a task), then changing the trigger and teaching a replacement response that gets the learner the same thing, is expected to reduce the behavior (behavioural outcome) more than a generic response chosen without that information, and more than escalating punishment. Triggers fall into three groups (Heineman, Dunlap & Kincaid, 2005, as the earlier page summarised them): physiological (illness, fatigue, hunger, medication), physical or environmental (heat, cold, seating, not being able to see or hear), and instructional (too little choice, unclear instructions, work too hard or too long, no way to ask for help). The second half of the relationship, from the earlier page: responses that are fair, consistent, documented and agreed with the family and other professionals make long-term support workable, and escalating punishment out of frustration tends to backfire.

The relationship is conditional on the behavior recurring often enough to observe (a rare or dangerous behavior cannot wait for a pattern), on the trigger being specific (behavior that occurs everywhere and all day points to a skill gap or a wider need, not a trigger), on someone having time to record and review, and on the adults carrying out the plan as written. Whether gains last after support is faded is a separate question the evidence here does not settle.

No claim in the wiki tests a function-based plan against a generic one, or tests documentation and consistency as such. The nearest evidence is: one functional analysis that identified escape and attention as what kept up one child's behavior (a single case, in a feeding programme); a meta-analysis showing that universal classroom management, chosen without functional information, reduces disruptive behavior substantially; single-case studies of self-monitoring; a randomised study of an antecedent change (greeting students at the door); and randomised studies of an empathic, non-punitive discipline mindset for teachers. Each is described below with its limits.

How this page sits beside its neighbours: the Positive Behavioral Interventions and Supports pattern places this assessment as the individual, most intensive tier of a schoolwide system, after universal practices. The strategy pages Functional Behavior Assessment and Functional Behavioral Assessment give the step-by-step procedure (interviews, ABC recording, experimental analysis). Self-Monitoring, a converted principle, holds learners checking their own on-task behavior; this page uses it late in a plan, as a way of handing control to the learner. The reinforcement ideas used here (reinforce the replacement, fade the reinforcer, distinguish a consequence from a reinforcer that measurably changes behavior) are set out on Behaviorism, which now holds the former reinforcement-theory principle.

Default design, while the relationship is untested

The earlier page's guidance, made concrete as a plan a teacher or support team can act on and revise. Each step is a design proposal with its evidence status beside it. Quantities are untested proposals unless a claim is cited.

  1. Check universal practice first. Before an individual assessment, confirm the class already has clear routines, taught expectations and frequent positive contact; universal classroom management reduces disruptive behavior for most students without any functional information (universal classroom management [~M]). Start an individual assessment when one learner's behavior persists for about two to three weeks despite it (untested threshold).
  2. Define one behavior in observable terms. "Calls out during independent work", not "is disrespectful". One behavior at a time, with a count, duration or intensity that two adults would record the same way. Untested here; from the strategy pages.
  3. Record trigger, behavior and what follows, across settings (week 1 to 2). Use a short antecedent–behavior–consequence (ABC) sheet: time, activity, what happened just before, the behavior, what adults and peers did just after, and what the learner got or avoided. Aim for about 10 to 15 recorded incidents across at least three activities or times of day, and also note times the behavior did not occur. Use the three trigger groups as a checklist: physiological (was the learner tired, hungry, unwell, newly on medication?), environmental (temperature, seat, noise, sightline), instructional (task length and difficulty, clarity of instructions, any choice offered, any way to ask for help). Ask the learner and the family the same questions. Untested proposal; the trigger groups are from the earlier page.
  4. Write a hypothesis with two parts: "When [trigger], [learner] does [behavior], and gets [attention / escape from the task / an object / sensory relief]." The evidence that functional analysis can separate these: one child's mealtime behavior was shown to be kept up by both escape and attention, and that result set the treatment (escape and attention [+W]; one child, a feeding programme, not a classroom). Expect more than one function.
  5. Test the hypothesis cheaply (week 2 to 3). For three to five sessions, change the suspected trigger (shorter task, instructions given one step at a time, a choice of two tasks, a seat nearer the board, a snack before the lesson) and keep recording. If the behavior drops, the hypothesis gains support; if not, return to step 3 rather than adding consequences. Untested proposal.
  6. Prevent: change the trigger where you can. Keep the change that worked. Antecedent changes that cost little can be built into the day: a brief, personal greeting at the door raised engaged time and reduced disruptive behavior in middle-school classes (greetings at the door [+M]; a randomised study of teachers and a three-student single-case study).
  7. Teach a replacement response that gets the same thing (weeks 3 to 6). If the behavior gets help or escape from a hard task, teach a help card or a "break, then back" request; if it gets attention, teach an appropriate way to get it and give attention for that. Model it, rehearse it twice in a calm moment, and prompt it just before the known trigger. Reinforce the replacement every time at first (specific praise, the help or break it asks for), and keep reinforcement for the old behavior as low as safety allows. Untested here; from the earlier page and the strategy pages (Behaviorism sets out reinforcement and fading). Prefer praise and the natural payoff over tangible prizes for activities the learner already likes: expected tangible rewards reduced later free-choice engagement in laboratory experiments, while verbal praise did not (expected tangible rewards [~M]).
  8. Respond to the incidents that still happen consistently, without escalating. Decide in advance what the adult does, keep it the same each time, and do not raise the "volume" of punishment when frustrated. A related finding: a short exercise encouraging teachers to treat misbehavior with an empathic rather than punitive mindset reduced suspensions, most for students previously suspended, though the preregistered overall effect was small and marginal (empathic discipline [+M]; middle-school math teachers; suspension is the outcome, not classroom behavior).
  9. Document both sides, every day, briefly. Record the behavior count and the adult's own response, so that the reasonableness of rules and reactions can be shown to the family and colleagues. Share a one-page summary weekly with the family and the team. Untested; from the earlier page. Where the plan is part of an individualised education programme (IEP), do not depart from it alone: a significant change needs the team and the family, which is a legal requirement in that setting, not an evidence claim.
  10. Hand control to the learner (weeks 6 to 10). Thin reinforcement for the replacement from every time to roughly every second or third time, then to the end of the lesson. For learners in about grade 4 and above, add self-recording: at set prompts the learner marks whether they were on task (self-monitoring of on-task behavior [+M]; single-case studies of students with or at risk for disabilities; on-task behavior, not learning). Untested as a schedule.
  11. Review against the baseline every two weeks. Compare counts with the first two weeks. Move to the next behavior, or to lighter monitoring, once the behavior has stayed at or below the agreed level for two consecutive weeks (untested threshold). If there is no change after about three weeks of the plan as written, check first whether it was carried out as written (low-fidelity implementation of positive behavior support went with smaller improvements: fidelity [~W]), then revisit the hypothesis. Do not respond to no change by escalating consequences.

Dose and progression, in summary (untested proposal): weeks 1 to 2, record and interview, no new plan; weeks 2 to 3, test one trigger change for three to five sessions; weeks 3 to 6, trigger change kept, replacement taught and reinforced every time, adult prompts before each known trigger; weeks 6 to 10, reinforcement thinned, adult prompts replaced by the learner's own record and a daily check with one adult; from then on, a weekly check. Each ABC entry takes under a minute; each review about ten minutes.

When a condition is not met: if the behavior is dangerous, do not wait two weeks to observe; put a safety plan in place and involve a behavior specialist at once. If the behavior occurs in every setting and activity, look for a skill the learner lacks (they do not know how to join a group, ask for help, or wait), and teach it directly (see Explicit Interpersonal Skills Instruction), or for a wider need that calls for referral.

Fitting the design to a situation

The three facts that most change the decision:

  • How often, where and how dangerously the behavior occurs. Frequent behavior in some settings and not others suits recording and a trigger hypothesis; rare or dangerous behavior needs a safety plan and specialist help first; behavior everywhere suggests a skill gap or a wider need. If unstated, ask: how many times a day or week, in which lessons or places, and has anyone been hurt?
  • Who can observe and record, and for how long. The plan above assumes an adult can make a one-minute note after each incident for two weeks and meet for ten minutes every two weeks. If unstated, ask: who sees the learner during the problem times, and how much time a week can they give to recording and review?
  • What the plan is for, and who must agree to it. Fewer disruptions in one lesson, staying in school (avoiding exclusion), or the learner managing their own behavior in a new setting need different targets and different checks; a formal plan (IEP or behavior plan) sets who must agree to changes. If unstated, ask: what would count as success, by when, and is there an existing written plan?
If the brief says… Then change… Basis
The whole class is disruptive, not one learner Do not start individual assessments; put universal management in place first (taught routines, clear expectations, frequent positive contact), then reassess which learners still stand out Universal classroom management [~M] (12-study meta-analysis, K–12 classrooms; universal practices without functional information)
Learner is a young child (preschool or early primary) Rely on adult observation and family interview, not the child's report; use a picture card or gesture as the replacement response; keep consequences immediate and the reinforcer frequent; check sleep, hunger and illness first untested proposal; the only functional-analysis claim here is one 4-year-old in a feeding programme (escape and attention [+W])
Learner is an adolescent (about grade 4 and up) Interview the learner as a main source of trigger information; agree the replacement with them; add self-recording of on-task behavior from week 4 or so Self-monitoring of on-task behavior [+M] (single-case studies, grades 4–12, students with or at risk for disabilities; behavior, not achievement)
Learners are adults in a work setting or adult education Drop the IEP and parent steps; ask the person directly what makes the session hard (shift fatigue, a task they cannot read, being put on the spot); agree the change with them privately; document only what the organisation's policy requires untested proposal
The behavior is dangerous, or rare but severe Do not wait to observe a pattern; put a safety plan in place and refer to a behavior specialist; record each incident in full untested proposal; from the strategy pages (rare behavior cannot be sampled)
One teacher with a full class, no support staff Use a tally sheet with three columns (time, activity just before, what followed) during the two lessons where the behavior is worst, for one week; test one trigger change in the second week untested proposal
Scarce adult time (volunteers, a busy shift, one tutor checking in briefly) Cut recording to a three-question card after each incident (when, what just before, what the learner got), and replace the two-weekly meeting with a five-minute weekly look at the cards by one named person untested proposal
Only one session (a cover teacher, a one-off workshop) Do not attempt an assessment; use antecedent defaults: greet each learner by name at the door, give instructions one step at a time, show a way to ask for help or take a short break, and pass a note on what happened to the regular teacher Greetings at the door [+M] (middle-school classes, carried to a single session it did not test); the rest untested
Mixed modes (some learners remote, some in the room; or phone-only) Add connection, camera, noise at home and screen size to the environmental trigger list; record from chat and platform logs as well as observation; give remote learners a private channel (direct message) as the help or break request untested proposal
The learner has an IEP or a behavior plan Keep the existing plan; record against it; bring proposed changes to the team and the family rather than changing it alone untested proposal; the plan is an agreement among teacher, professionals, learner and parents (earlier page), and US guidance holds that changes to an IEP go through the IEP process (IEP processes protect due process [+W], asserted guidance, not a finding)
The brief names its own routine (a point sheet, a daily behavior report card, check-in/check-out, a class reward chart) Keep the routine; add one functional check (does the reward or attention the routine gives match what the behavior was getting?) and change only what the check shows, for example the timing of the check-in to just before the known trigger untested proposal; see Check In Check Out and Daily Behavior Report Cards
The plan is to reward the learner with prizes for an activity they already enjoy Use specific praise and the natural payoff (help, a break, time on a chosen task) instead; if tangible rewards are needed, tie them to meeting a standard and fade them Expected tangible rewards [~M] (two meta-analyses of experiments, mostly laboratory tasks; they disagree on size; stronger for children)
Staff or the school want harsher consequences or exclusion Present the record and the hypothesis; agree consistent, pre-planned consequences; offer teachers a short empathic-discipline exercise Empathic discipline [+M] (middle-school math teachers; suspensions; overall effect small and marginal in the preregistered replication, larger for previously suspended students)
Learners with low literacy, a language barrier or limited communication Make the replacement response non-verbal (a card, a gesture, a signal); interview through the family or an interpreter; treat unclear instructions as a likely instructional trigger untested proposal
The behavior occurs when the learner must perform for others (reading aloud, presenting, a public performance) Hypothesise escape from being seen to fail; offer a rehearsal with one adult first, then a small group, then the audience; allow an alternative format while that builds untested proposal
Staff carrying out the plan have little training Coach the adults on the plan in the classroom and check, at each review, whether each step was done as written before judging the plan Low fidelity [~W] and teacher fidelity [~M] (one study, 37 students with emotional disturbance in self-contained classrooms, no control group)

Seven of the sixteen rows cite a claim, and none of those claims tests a function-based plan. The universal-management, greetings and empathic-discipline rows carry findings about classroom-wide or teacher-level practice to an individual plan; the self-monitoring row comes from single-case studies with students with or at risk for disabilities; the rewards row from laboratory experiments; the fidelity rows from one uncontrolled study; the young-child row cites its claim only to show how thin the evidence for children is. The rest are untested proposals.

Observation, state and explanation

What can be observed is a behavior in a setting: its time, the activity, what happened just before, what adults and peers did just after, what the learner got or avoided, and how often it occurs. "Attention-seeking", "avoidant", "defiant" and "lacks the skill" are inferred states, and the same behavior can serve different functions for different learners, or for one learner at different times.

Response Plausible explanations Observation that could change the interpretation
Calls out or disrupts during independent work Gets adult attention; the work is too hard and the disruption gets help or escape; the learner does not know how to ask for help Give the same work at an easier level for two sessions; separately, give attention every few minutes regardless of behavior. Which one lowers the rate?
Refuses or leaves when a task is set Escape from task difficulty or length; tired, hungry or unwell at that time; no choice offered Compare the same task in the morning and afternoon, a shorter version, and a choice of two tasks; ask the learner what makes it hard.
Behavior clusters at one time of day or in one room A physiological pattern (before lunch, after medication); an environmental one (heat, noise, seat); the subject or the adult in that lesson Move the activity to another time or room for a week; check medication and meal times with the family.
Behavior falls, then returns Reinforcement thinned too fast; a new trigger; adults have drifted from the plan; the replacement stopped working for the learner Check the record of adult responses against the plan, then restore the earlier reinforcement level for a week.
Behavior occurs in every setting, with every adult A skill the learner lacks; a wider need (anxiety, trauma, an unmet learning need) rather than a specific trigger Teach the missing skill directly and see whether the behavior falls; if not, refer for a wider assessment.
Behavior gets worse when a consequence is applied The consequence removes the learner from the task (escape) or gives attention, so it reinforces the behavior Record what the learner gets from each consequence for a week; replace consequences that give escape or attention.

These rows are proposals, untested as diagnostics. Changing a condition to test a hypothesis also changes the behavior, so record each test and its dates, and keep two explanations open while both still fit the record.

Evidence and qualifications

No claim in the wiki compares a function-based plan with a plan chosen without functional information, or tests trigger identification, documentation or consistency of response as such. The claims below bear on parts of the relationship.

  • Inappropriate mealtime behavior was maintained by escape and attention [+W]: in a day-treatment feeding programme, a functional analysis with escape, attention and control conditions showed that one child's inappropriate mealtime behavior was kept up by both escape and attention, which set the baseline and the extinction treatment (Groff et al. 2011; one participant; coded q2, causal r1; no effect size). It shows a functional analysis separating two functions in one case, and that one behavior can serve two. It is not a classroom study and does not compare function-based with other treatment.
  • Teachers' universal classroom management practices significantly reduce problem classroom behavior [~M]: a random-effects meta-analysis of 12 studies of whole-classroom management interventions in K–12 classrooms (Oliver, Wehby & Reschly 2011) reports a mean classroom effect of 0.80 (ICC = .05) or .71 (ICC = .10), ranging from −0.05 to 1.74, coded q3 i2. It qualifies this page: classroom-wide practice, chosen with no functional information about any one student, reduces disruption substantially, so the earlier page's contrast with "a generic behavior-management approach" holds only for the learner whose behavior persists despite it. The same claim records that students in treatment classrooms were less disruptive in all 12 studies, and a sibling claim from the same review, effect sizes across the studies were homogeneous, so no moderator analyses were justified [~M], which means the review cannot say for which students or settings universal practice is enough.
  • Positive Greetings At The Door Improve Engagement [+M]: middle-school teachers randomly assigned to greet students individually at the door, against an attention control, produced more academic engaged time and less disruptive behavior (Cook et al. 2018, q3); in a three-student multiple-baseline study, on-task behavior rose from a mean of 45% to 72% (Allday & Pakurar 2007, q1). Neither reports a standardised effect size. It supports changing what comes before a behavior as a cheap preventive move; it is a class-wide antecedent, not one chosen from an assessment.
  • Self Monitoring Improves On Task Behavior [+M]: two meta-analyses of single-case studies with students with or at risk for disabilities (grades 4–12 in one) found on-task behavior rose when students recorded their own behavior at prompted intervals; in one, effects were similar with or without added reinforcement. Effects are reported as response ratios and overlap statistics, not standardised effect sizes, and the outcome is on-task behavior, not learning. It supports self-recording as the later stage of a plan.
  • A brief online empathic-discipline exercise for teachers halved suspensions in one randomized field experiment, while its preregistered replication found a small, marginal overall effect but narrowed group disparities [+M]: a brief online exercise encouraging middle-school math teachers to take an empathic rather than punitive approach to discipline halved year-long suspension rates (9.6% to 4.8%, 31 teachers; abstract read); a preregistered, active-controlled replication (66 randomised teachers, 5,822 students) found a small, marginal overall effect and larger reductions for Black and Hispanic students and previously suspended students. It bears on the earlier page's warning against escalating punishment; the outcome is suspension, and the intervention targets teachers' mindset, not an individual plan. The claim's evidence would allow a stronger marker; it is held at M because it is carried from a teacher-mindset intervention to this page's individual plans.
  • Expected tangible rewards reduce later free-choice engagement with a rewarded task, while verbal praise does not [~M]: two meta-analyses of experiments (128 and 96 studies) disagree on the size of the undermining effect but agree that verbal praise does not undermine intrinsic motivation; one found tangible rewards more detrimental for children than for college students. It qualifies how a replacement response is reinforced; the studies are mostly laboratory tasks, not behavior plans.
  • Teachers' PBIS knowledge and fidelity (TKSS score) interacts with pre-to-post change in problem behavior among students with emotional disturbance [~M], which also records the authors' reading that teachers implementing PBIS with low fidelity tended not to see large behavior improvements: one pre–post study of 37 students with emotional disturbance in self-contained classrooms after a year of teacher training in positive behavior support (Benner et al. 2008), with no control group; the low-fidelity claim is the authors' interpretation (q1). It supports checking whether a plan was carried out before judging it.

Learners, settings and outcomes differ across these claims (one child in a feeding programme, whole K–12 classrooms, students with disabilities in single-case designs, middle-school suspension records), so do not rank them by effect label. None measures how long a reduction in behavior lasts after a function-based plan is faded.

The evidence by question

Further evidence, not yet read against this model

The earlier page cited no claims. These claims bear on neighbouring parts of the relationship and were found while converting it; none has been read against the model above beyond its own claim page.

Objective and learner-valued goal

The designer's or school's objective is usually fewer incidents of one behavior in stated settings, measured against a baseline, by a stated date, and sometimes the learner staying in class rather than being excluded. Elicit separately what the learner wants: to avoid looking unable in front of peers, to get help without asking out loud, to have a break when overwhelmed, to be treated fairly, to stay with friends. The function of the behavior often is a goal the learner values, met in a way that costs them, so agreement can be built in: a replacement response that meets the same need (a private help signal for a learner who values not looking stuck) serves both. Divergence is an observation too: a learner who values leaving the room will not keep a plan whose only payoff is staying.

For example, a Year 8 student who walks out of mathematics may value not being seen to fail; the teacher's objective is that the student completes the lesson. A plan that removes public answering, adds a help card, and starts each lesson with an item the student can do serves both. Record the student's goal and the objective separately, so that "stayed in the room" is not read as "learned the mathematics".

What would revise this model?

The expectation should weaken if studies comparing function-based plans with plans chosen without functional information, in ordinary classrooms, show no difference in behavior at follow-up; or if brief classroom recording by teachers turns out to agree poorly with what an experimental analysis finds, in which case the default design needs a specialist step rather than a teacher-run one. If universal classroom management, done well, leaves almost no learners with persisting behavior, individual assessment is a rare specialist tool rather than a routine one. If consistent, documented responses make no difference to family collaboration or outcomes, the second half of the earlier page's relationship should be dropped. Do not protect the model by calling every failure "the wrong function" after the fact; record the hypothesis and the test before the plan starts.

A reduction in behavior during the plan, its maintenance after fading, its spread to other settings, and any effect on learning are separate claims. The present evidence establishes no recording period, no number of observations, no reinforcement schedule and no criterion for ending a plan.

  • Explicit Interpersonal Skills Instruction — a complementary strategy for students whose disruptive behavior stems from an actual social-skills gap rather than a specific environmental trigger
  • Self-Monitoring — learners checking and recording their own on-task behavior; used here as the later, learner-run stage of a plan
  • Clear Structure — clear routines and instructions, the universal practice that comes before individual assessment and removes one common instructional trigger
  • Behaviorism — the theory page holding reinforcement, extinction and fading (and the former reinforcement-theory principle) that a function-based plan uses

Examples

Key Sources

  • Heineman, M., Dunlap, G., & Kincaid, D. (2005). Positive support strategies for students with behavioral disorders in general education settings. Psychology in the Schools, 42(8), 779-794.
  • Arduini-Van Hoose, N. (2020). Behavior disorders. In Educational psychology. Retrieved from https://edpsych.pressbooks.sunycreate.cloud. CC BY-NC-SA 4.0.
Page metadata
Field Value
type principle
id functional-behavior-assessment
title Functional Behavior Assessment
description For a learner whose disruptive behavior recurs despite sound classroom management, identifying what triggers it and what it gets them, then changing the trigger and teaching a replacement that meets the same need, with consistent documented responses, is expected to reduce it more than generic or escalating discipline, though no claim here tests a function-based plan against a generic one.
status review
canonical True
generated by claude/unspecified on 2026-10-05
trust unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.