📖12 min read · 2,400 words
This is the plain-language version. Same content, written for a quick read.
ABC narrative recording is the most-used kind of behavior data, and the most often poorly written. A good ABC sample gives the board certified behavior analyst (BCBA) or supervising teacher solid raw material for figuring out why a behavior is happening. A poor one is just paperwork no one can use. The effort is about the same. The value is very different.
This brief covers what ABC means, what each column should and should not hold, how to take notes in real time, common errors, and what happens to the notes after you write them.
ABC is not the only kind of behavior data. See brief 06.01 for the bigger list (how often, how long, how soon, prompt level, work products, interval notes). ABC works best for puzzling, escalated, or rare events where the setting matters and counts won't tell you much. For known patterns that happen a lot, frequency or interval notes are usually more helpful.
Three columns:
Antecedent (A): what was happening right before the behavior. The setting, who was there, what was being asked, what the student had been doing. The trigger, in plain words.
Behavior (B): what the student did. What you saw and heard. Specific, not what you think it means.
Consequence (C): what happened right after. What the adult did, what other kids did, what changed in the room, what the student got or got out of.
ABC notes work because of the function idea underneath them: behavior is shaped by what comes before and what comes after. A pattern of A's that keep showing up before the behavior, and C's that keep showing up after it, points to function. (See brief 05.01 on function-based thinking.)
Should hold:
Should NOT hold:
Should hold:
Should NOT hold:
Should hold:
Should NOT hold:
ABC notes can be in full sentences or in tables. Tables are usually quicker to scan and easier to keep up with in real time. Apps and digital systems are out there (BehaviorSnap, Catalyst, Rethink); paper still works fine.
| Field | Example entry |
|---|---|
| Date / time | 5/8 9:32 a.m. |
| Setting | 3rd grade math, room 207, whole-group lesson |
| Antecedent | Teacher passed out worksheet on multi-digit subtraction. Arjun sighed out loud. Teacher said 'You can do this, just try the first one.' |
| Behavior | Arjun pushed worksheet off desk, said 'No,' put head down on desk, hood up. Lasted 4 minutes. |
| Consequence | Para came over, sat in the next chair, said 'Take a minute, I'll be right here.' Did not make him do the worksheet. After 4 min, Arjun picked up worksheet and started problem 1 with verbal prompts. Worksheet done at 9:48 with 50% right. |
| Notes / guess | Maybe escape from a task he doesn't like. Setting event: parent said he didn't sleep much last night. |
"5/8, 9:32, math. Worksheet given; M sighed; teacher said 'try it.' M pushed paper, head down 4 min. Para sat nearby, did not require work. M started after break with verbal prompts."
This packs the same info into one short paragraph. Useful when you have to write down an event on the fly and will fill in detail later.
Some paras carry a small notebook or sticky pad in a pocket. Write A/B/C as bullet points or three short lines on a sticky in the moment, then move it to the formal sheet at the end of the period or end of day. Memory fades within hours; even a few words written in the moment beat a polished sentence written from memory.
ABC notes are most useful when written within minutes of the event. Writing them at end of day from memory adds drift: antecedents get blurry, the order gets shuffled, words students said become words you remember them saying.
During a crisis, your job is the student's safety, not the paperwork. Catch the time on a clock when you can; fill it in as soon as you have a moment. Some teams have one para focused on the student and another writing notes: the team's plan for who does what matters more than always being able to write while supporting.
| Error | Why it matters / how to fix |
|---|---|
| Guesses in the B column | "Got frustrated" is a guess. "Pushed worksheet, head down, hood up" is what you saw. Save the guess for a hypothesis line; keep B to what you saw. |
| Vague time markers | "Earlier this morning" can't be used. "9:32" can. Use clock times. |
| Missing the C | Many ABCs trail off in the C column. The C is the most useful column: it's the function clue. Write it fully. |
| Bias toward big events | If you only write ABC for crisis moments, you miss the everyday moments where the same function shows up at lower intensity. Sample some calm-but-hard moments too. |
| Bias against the student | Quoting the adult word for word while paraphrasing the student, or the other way around, tilts the record. Quote both sides word for word: adult words and student words, plus what each of them did. |
| Not noting setting events | If the student didn't sleep last night, that belongs in the notes. Setting events explain why a trigger that usually doesn't matter went sideways today. |
| Writing one ABC and stopping | ABC is a sampling tool. Patterns show up across many entries. One event is a story; ten events are data. |
| Editing to be kind | Softening the student's words or actions to spare their feelings makes the notes useless. Be accurate; a kind team will still read it kindly. |
| Writing what the BIP says will happen rather than what did happen | If the BIP says "give a break card" and you actually said "sit back down," write what you did. The team can't fix problems they don't see. |
ABC notes feed the team's thinking. They are most useful when:
| Use ABC when… | Use other formats when… |
|---|---|
| You're trying to find the function (FBA-prep). | The function is already known and you're tracking how often. |
| A behavior is rare or high-stakes (rare aggression, elopement, leaving the room or building without permission, or serious self-injury). | A behavior happens a lot and you need counts (frequency, interval recording). |
| A pattern is unclear and you need the setting. | You're tracking prompt level for a teaching skill (prompt-level data, see 06.03). |
| A new behavior has shown up and you don't know why. | You're keeping work samples. |
| The team is thinking about doing an FBA. | The student is on a steady BIP and you're checking how well it's being run. |
Three poorly-written entries:
None of these is usable. No times. No specific antecedents. Behavior is labeled, not described. Consequences are vague. The team can read these for a year and learn nothing about function.
Three well-written entries:
These are usable. A reader can picture the event. Antecedents are specific. Behavior is described, not labeled. Consequences include both what the adult did and what happened next. The team can look at function across many entries and sharpen the BIP.
How to take Antecedent–Behavior–Consequence notes that the team can actually use
Why this brief
ABC narratives are the simplest, most-used form of behavior data, and the most often poorly written. A well-taken ABC sample is the BCBA's or supervising teacher's best raw material for understanding why a behavior is happening; a poorly-taken one is descriptive paperwork that no one can analyze. The difference is small in effort and large in usefulness.
This brief covers what ABC means, what each column should and should not contain, formats for taking notes in real time, common errors, and what happens to the data after you take it.
1\. What ABC means
Three columns:
Antecedent (A): what was happening immediately before the behavior. The setting, who was there, what was being asked, what the student had been doing. The trigger, in plain language.
Behavior (B): what the student did. Observable, descriptive, specific. Not interpretation.
Consequence (C): what happened immediately after. Adult response, peer response, environmental change, what the student got or got out of.
ABC notes work because of the function-based logic underneath them: behavior is shaped by what comes before and what comes after. A pattern of A's that reliably precede the behavior, and C's that reliably follow it, points to function. (See brief 05.01 on function-based thinking.)
2\. What each column should (and should not) contain
2.1 Antecedent
Should contain
Time and date.
Setting (location, period, activity).
Who was present (which adults, which peers).
What was being asked of the student or what they were doing.
Anything in the immediate environment that might be relevant (noise, transitions, schedule changes).
Setting events you know about (slept poorly, missed meds, fight at home): note these even if they happened earlier in the day.
Should NOT contain
Interpretation. "He was getting frustrated" is not antecedent: it's already a hypothesis. The antecedent is what set up the frustration: "asked to start writing."
Mind-reading. "She didn't want to do the work": not observable. "Was given a math worksheet at 9:32; turned away; pushed paper to the floor": observable.
Generalities. "Earlier today" is not enough; pin down what the student was doing 30 seconds before.
2.2 Behavior
Should contain
What the student actually did. Words used (in quotes), motor behavior, voice, body.
Topography: the form of the behavior. "Pushed worksheet off the desk and onto the floor; pulled hood up over head; said 'leave me alone' in a quiet voice; head down on desk."
Duration: how long it lasted, if relevant.
Intensity: for behaviors that vary in intensity. "Voice rose to a yell. Stomped feet. Threw chair across the room" reads very differently than "voice rose to a yell. Stomped feet. Walked to the cubby."
Should NOT contain
Inference about why. The B column describes; the A and C columns are where the why-pattern emerges across many entries.
Labels. "Tantrum" or "meltdown" or "aggression" can be a useful tag, but the description has to be specific enough that someone reading it can picture the moment.
Adult emotion or judgment. "Was very disrespectful" tells me how you felt; not what the student did.
2.3 Consequence
Should contain
What adults did, exactly. Words used (in quotes), procedures invoked, who came over, what was offered, what was withheld, where the student ended up.
What peers did. "Class went quiet for 10 seconds, then resumed." "Two students laughed."
Environmental changes. "Bell rang." "Lunch arrived." "Lights flickered."
What the student got, or got out of. The escape ("sent to hallway, no longer doing math"), the attention ("adult sat down beside her"), the tangible ("received iPad"), the sensory ("removed jacket and stretched"). This is the heart of the function analysis.
Should NOT contain
What you wish had happened. Document what did.
Justifications. "I had to send him out because…": note the action, not the rationale.
Future plans. "Going to talk to mom": that goes in a separate field, not in the C column.
3\. Practical formats
ABC narratives can be in written sentences or in tables. Tables are usually faster to scan and easier to keep up with in real time. Apps and digital systems exist (BehaviorSnap, Catalyst, Rethink); paper still works fine.
3.1 A simple ABC table format
| Field | Example entry |
|---|---|
| Date / time | 5/8 9:32 a.m. |
| Setting | 3rd grade math, room 207, whole-group instruction |
| Antecedent | Teacher distributed worksheet on multi-digit subtraction. Arjun sighed audibly. Teacher said 'You can do this, just try the first one.' |
| Behavior | Arjun pushed worksheet off desk, said 'No,' put head down on desk, hood up. Lasted 4 minutes. |
| Consequence | Para approached, sat in adjacent chair, said 'Take a minute, I'll be right here.' Did not require worksheet. After 4 min, Arjun picked up worksheet and started problem 1 with verbal prompts. Worksheet completed at 9:48 with 50% accuracy. |
| Notes / hypothesis | Possibly escape from non-preferred task. Setting event: parent reported limited sleep last night. |
3.2 Quick narrative format for short entries
"5/8, 9:32, math. Worksheet given; M sighed; teacher said 'try it.' M pushed paper, head down 4 min. Para sat nearby, did not require work. M started after break with verbal prompts."
This compresses the same information into one paragraph. Useful when you have to capture an incident on the fly and will fill in detail later.
3.3 Sticky-note method for paras in motion
Some paras carry a small notebook or sticky pad in a pocket. Capture A/B/C as bullet points or three short lines on a sticky in the moment, then transfer to the formal sheet at the end of the period or end of day. Memory fades within hours; even a few words written in the moment beats a polished sentence written from memory.
4\. Taking ABC in real time
ABC notes are most useful when written within minutes of the incident. Writing them at end of day from memory introduces drift: antecedents get blurry, sequences get reshuffled, words students said become words you remember them saying.
4.1 In the moment
Note the time.
Note who was present and what was happening before.
Note exactly what the student did, in their words and movements.
Note what happened next.
Don't try to write a paragraph; write notes you can expand later.
4.2 When you can't write in the moment
During a crisis, your job is the student's safety, not the paperwork. Capture the time on a clock when you can; reconstruct as soon as you have a moment. Some teams have one para focused on the student and another writing notes: the team's agreed-upon division of labor matters more than always being able to write while supporting.
5\. Common errors
| Error | Why it matters / how to fix |
|---|---|
| Interpretation in the B column | "Got frustrated" is interpretation. "Pushed worksheet, head down, hood up" is observation. Save the interpretation for a hypothesis line; keep B observable. |
| Vague time markers | "Earlier this morning" can't be analyzed. "9:32" can. Use clock times. |
| Missing the C | Many ABCs trail off in the C column. The C is the most analytically valuable column: it's the function clue. Write it fully. |
| Bias toward dramatic incidents | If you only write ABC for crisis episodes, you miss the everyday moments where the same function shows up at lower intensity. Sample some calm-but-difficult moments too. |
| Bias against student | Quoting the adult verbatim while paraphrasing the student, or the reverse, slants the record. Quote both sides verbatim: adult words and student words, plus what each of them did. |
| Not noting setting events | If the student didn't sleep last night, that belongs in the record. Setting events explain why a normally-tolerated trigger went sideways today. |
| Writing one ABC and stopping | ABC is a sampling tool. Patterns emerge across many entries. One incident is a story; ten incidents are data. |
| Editing for kindness | Softening the student's words or actions to spare their feelings makes the record useless. Be accurate; a sympathetic team will still read it sympathetically. |
| Writing what the BIP says will happen rather than what did happen | If the BIP says "give a break card" and you actually said "sit back down," record what you did. The team can't fix fidelity issues you don't surface. |
6\. What happens to the data
ABC notes feed the team's analysis. They are most useful when:
There is enough volume to see patterns (typically 10+ incidents over 1–2 weeks).
Multiple writers contribute (different paras, the supervising teacher, the gen-ed teacher): different perspectives catch different things. Those writers still have to be calibrated against each other, or differing perspective becomes differing standards and the entries stop being comparable. Have the team write up the same incident independently, then compare column by column with the supervising teacher.
Someone is reviewing them. ABC notes that no one reads might as well not exist. Confirm with the supervising teacher who reviews and how often.
The team uses them to refine hypotheses, antecedent strategies, or BIP procedures: not to assign blame.
6.1 When to take ABC vs. when to take other data
| Use ABC when… | Use other formats when… |
|---|---|
| You're trying to identify function (FBA-prep). | Function is already established and you're tracking frequency. |
| A behavior is low-frequency or high-stakes (rare aggression, elopement, severe self-injury). | A behavior is high-frequency and you need counts (frequency, interval recording). |
| A pattern is unclear and you need context. | You're tracking prompt level for an instructional skill (prompt-level data, see 06.03). |
| A new behavior has appeared and you don't know why. | You're documenting permanent products of student work. |
| The team is considering whether to do an FBA. | The student is on a stable BIP and you're monitoring fidelity. |
7\. A worked example
Three poorly-written entries
"Arjun had a meltdown in math today. He just refused everything. Got him calmed down eventually."
"Priya was being defiant during reading. She wouldn't do the assignment. I had to send her out."
"Aggressive episode at lunch. Talked to him about better choices."
None of these is usable. No times. No specific antecedents. Behavior is labeled, not described. Consequences are vague. The team can read these for a year and learn nothing about function.
Three well-written entries
"5/8 9:32. 3rd grade math, whole group. Teacher distributed double-digit subtraction worksheet. Arjun sighed audibly. Teacher said, 'You can do this, just try the first one.' Arjun said, 'No,' pushed worksheet to floor, head down, hood up. Para sat in adjacent chair, said, 'Take a minute, I'll be here.' Did not require worksheet. After \~4 min, Arjun picked up worksheet, started problem 1 with verbal prompts. Worksheet completed by 9:48, \~50% accuracy. Setting event noted: parent reported only 5 hours sleep."
"5/8 11:14. 3rd grade reading, small group of 4. Priya asked to read aloud. Priya looked at adjacent peer, looked down, said nothing. Teacher said, 'I'll wait.' 12-second silence. Teacher said, 'OK, let's try together.' Priya said, 'I'm not doing it,' stood up, walked toward door. Para said, 'Priya, you can sit with me at the back table.' Priya walked to back table, sat with head on table. Did not return to group. Read with para 1:1 at 11:30 with no resistance, completing assigned passage at instructional level."
"5/8 12:18. Cafeteria, lunch. Arjun seated next to Kenji at usual table. Kenji reached toward Arjun's lunchbox. Arjun shouted 'no\!' and shoved Kenji's hand away. Kenji said 'sorry' and pulled hand back. Arjun continued eating. No further incident. Para reminded both about asking before reaching."
These are usable. A reader can picture the incident. Antecedents are specific. Behavior is described, not labeled. Consequences include both the adult response and what happened next. The team can analyze function across multiple entries and refine the BIP.
8\. Common pitfalls
Writing ABCs only for the worst behaviors. The same function often shows up in lower-intensity moments and is easier to learn from.
Writing from memory at end of day rather than noting in the moment.
Treating ABC as a punitive tool rather than an analytic one. The records exist to help the team understand, not to build a case.
Letting the form constrain the substance: if the form has too few lines for what you saw, write more.
Different paras using different conventions. Calibrate as a team.
Not noting your own response in the C column. Adult behavior is data.
Forgetting setting events. The night-before sleep, the morning conflict, the missed meds: these belong in the record.
Recording but not reviewing. ABC piles that no one reads are dead weight.
9\. Resources
PBIS; FBA and Behavior Documentation; pbis.org; Free national-level technical resources.
IRIS Center; Functional Behavior Assessment; iris.peabody.vanderbilt.edu; Free self-paced module.
O'Neill, Albin, Storey, Horner, Sprague; Functional Assessment and Program Development; Cengage; Foundational text on FBA, including ABC narrative methods.
BehaviorSnap, Catalyst, Rethink; vendor sites; Common digital ABC/data platforms; check FERPA and district approval before use.
Brief 05.01; Function-Based Thinking
Reading is useful, but recall is where it sticks. Three short scenarios, low-stakes, no scoring. About 3 minutes. You can stop any time.
Start the practice set →You take data on student goals — and you want to know which type (frequency, duration, latency, inte…
You're tracking behavior that's hard to count (engagement, vocal stim, on-task) — and you want to kn…
You're teaching a skill — and you want to track honestly how much help the student actually needed,…
You're collecting data that feeds quarterly IEP progress reports — and you want to be sure the repor…