What ABC data are
ABC stands for antecedent, behavior, and consequence. The antecedent is the event or condition immediately before the behavior - an instruction, a transition, a toy being taken away, an adult turning to someone else. The behavior is what the learner did, described in observable terms. The consequence is what immediately followed - attention, the removal of a task, access to an item, or nothing at all.
ABC recording is a form of descriptive assessment: it observes behavior as it occurs in everyday conditions, without changing anything to test a hypothesis. It is a standard part of a functional behavior assessment (FBA), alongside interviews and rating scales, and it is often paired with notes on setting events - broader conditions such as poor sleep, illness, or a change in routine that make a behavior more likely on a given day.
Three ways to collect descriptive data
ABC data can be written as free narrative or recorded against fixed categories, and a scatterplot is often run alongside either to show when the behavior happens. Most assessments start narrative and move to structured recording once the likely antecedents and consequences are known.
| Method | How it works | Best for | Watch for |
|---|---|---|---|
| Narrative ABC | The observer writes a free-text account of the antecedent, behavior, and consequence each time the behavior occurs. | Early in an assessment, before the likely triggers and outcomes are known. | Interpretive wording, inconsistent detail between observers, and records that are slow to tally. |
| Structured ABC | The observer selects from predefined antecedent and consequence categories, such as "demand placed" or "attention given", with room for a note. | Once the likely categories are known; quick to record and easy to count. | Events that fit no category get forced into one, and anything unlisted is lost. |
| Scatterplot | The observer marks whether the behavior occurred in each time block of the day, across many days. | Revealing time-of-day, activity, or staff patterns that individual records miss. | It shows when a behavior happens, not why; it needs ABC records alongside it. |
Writing an ABC record someone else can use
A good record describes what an observer saw and heard, not what they concluded. "He was frustrated" is an inference; "he pushed the worksheet off the table and said 'no'" is an observation another clinician can recognize, count, and compare. The behavior should match a written operational definition, so every observer is recording the same thing.
Record as soon as possible after the incident, with the time, setting, and people present, and note duration or intensity where they matter. The comparison below shows the same incident recorded both ways.
| Part | Interpretive | Observable |
|---|---|---|
| Antecedent | He got upset about having to work. | The clinician said "time to clean up" and took the tablet after 10 minutes of use. |
| Behavior | Had a meltdown. | Dropped to the floor, screamed, and kicked the table leg for about 45 seconds. |
| Consequence | Calmed him down. | The clinician sat beside him and talked to him for 2 minutes; the tablet was not returned and cleanup was delayed. |
How much ABC data is enough
There is no fixed number of records. What matters is whether the records cover the situations the behavior occurs in - different settings, activities, times of day, and people - and whether the pattern stays stable as new records arrive. Ten records from one cooperative afternoon say less than ten spread across a week of ordinary days.
Collect across observers, too. If only one person records, their habits of noticing shape the data. ABC categories can be checked for agreement like any other measure, and agreement on structured categories is a reasonable test of whether the definitions are clear enough to use; see interobserver agreement.
Reading ABC data for function
Behavior is commonly grouped into four functions: gaining attention, escaping or avoiding something, gaining access to an item or activity, and automatic reinforcement, where the behavior produces its own sensory consequence. ABC patterns suggest which function is likely. A behavior that follows demands and is followed by the demand being removed points toward escape. One that follows an adult's attention moving elsewhere and is followed by attention returning points toward attention. One that follows a preferred item being removed and is followed by getting it back points toward access. One that occurs across conditions, including when the learner is alone, with no consistent social consequence, points toward automatic reinforcement.
Count rather than eyeball. For each antecedent and consequence category, work out the share of incidents in which it appeared. Then remember what ABC data cannot see: they sample only the moments when the behavior occurred. If attention follows almost everything that happens in a one-to-one session, it will follow most incidents too, without being the reason for them.
What ABC data cannot tell you
ABC data are correlational. They can generate a strong hypothesis about function, but they cannot confirm one, because nothing in the environment was changed to test it. A functional analysis does that, by systematically arranging conditions and measuring the behavior in each; it is designed and supervised by a BCBA, and whether one is appropriate depends on the behavior's severity, the setting, and the risks involved.
Observation also changes what is observed. Staff who know they are recording may respond differently, and a learner may behave differently with an extra adult in the room. Treat ABC data as one source in an assessment, read alongside interviews, rating scales, and a quantitative measure of the behavior itself - frequency or duration, from the eight ABA measurement systems.
How Cadence records ABC data
ABC recording is one of the 8 measurement systems built into Cadence. Each incident is kept as its own record, with the antecedent, behavior, consequence, setting events, intensity, and time it occurred, and the session-level value is a count of incidents - so the narrative stays available for analysis while the target still has a graphable number.
ABC records usually run alongside a quantitative measure of the same behavior, and functional behavior assessment is supported as an assessment instrument, so the hypothesis the records feed lives with the assessment it belongs to.