Key takeaways
- Fill in the header once: behavior question, observable definition, and observation window.
- Each row needs a time, antecedent, behavior, and consequence; the other columns are optional.
- Log right after the window, not during it, and mark skipped sessions as skipped.
- After five sessions, tally the antecedent and consequence columns to see the pattern.
An ABC observation log is a table with one row per incident and a column each for the antecedent (what happened right before), the behavior, and the consequence (what happened right after). The worksheet below adds a short header and three optional columns, shows one filled row, and leaves blank rows so you can copy it and start logging tomorrow.
The header: fill it in once
Write these five fields at the top before the first session. They keep every observer pointed at the same behavior and make sessions comparable.
| Field | What to write | Example |
|---|---|---|
| Student and observer | Initials or a code, plus who is logging | Student J.R., observer: classroom teacher |
| Behavior question | The one thing the team wants to learn | What happens right before J.R. leaves the classroom? |
| Behavior definition | Observable wording two people would score the same way | Leaves the classroom door without permission |
| Observation window | A fixed block, so sessions compare | Literacy block, 9:00 to 9:30, Monday to Thursday |
| Setting events to watch | Background conditions the team already suspects | Missed breakfast, substitute, late bus |
The log
Four columns are required: time, antecedent, behavior, and consequence. Setting and activity, the intervention you tried, and setting events are optional. Keep them only if every observer will fill them every time.
| Date and time | Setting and activity | Antecedent (right before) | Behavior (as defined) | Consequence (right after) | Support tried and response | Setting event |
|---|---|---|---|---|---|---|
| Tue 9:12 | Classroom, independent writing | Writing prompt handed out, told to start alone | Left the room without asking, walked to the water fountain | Para walked him back; prompt paused 6 min, not finished | Offered break card on return; did not use it | Arrived late, no breakfast |
Notice what the filled row avoids. It does not say "got frustrated" (a guess about feelings) or "was defiant" (a label). It records the task, the action, what the adult did, and what happened to the task, which is the detail the team needs to see a function.
How to fill each column
- Date and time. Write the clock time, not "morning." Time of day is often the first pattern to show up.
- Antecedent. Name the event in the room right before the behavior: a demand, a transition, attention moving to another student, a preferred activity ending, or a change like a fire drill. If what you wrote is itself a behavior ("got angry"), back up one step.
- Behavior. Use the header definition, then add one concrete detail. Do not paraphrase the definition differently each time.
- Consequence. Record what happened right after, including the adult response, peer reactions, and what happened to the task. This column shows what keeps the behavior going.
- Support tried and response. If you used a planned support, note it and what the student did next. Leave it blank if nothing was tried.
- Setting event. Note unusual background conditions for the day, or leave it blank on an ordinary day.
Log right after the window closes, while it is fresh, rather than stopping instruction to write. If the window got away from you, write "session skipped" on the row. A skipped session is honest data; a reconstructed one is not.
Reading the log after five sessions
Once you have several sessions, read across the columns instead of down the rows.
Antecedent column: independent writing task (9 rows), transition to specials (3 rows), peer conflict (2 rows).
Consequence column: task paused or not finished (10 rows), adult one-on-one time (3 rows), no change (1 row).
Rows with the trigger present and no behavior: two partner-writing sessions, both with the student in seat.
Read: leaving follows solo written work and usually ends with the task delayed. That points toward escape from independent writing, which the team would test against the other functions before writing a hypothesis.
When the same antecedent and consequence repeat across settings and observers, you are ready to draft a statement. The guide to writing function hypothesis statements shows the format, and how much ABC data is enough covers when to stop collecting. If the log keeps dying by midweek, ABC data collection teachers will keep up with explains how to cut it down.
Moving the log into a plan
The log is raw material for the FBA summary and then the behavior plan. The practical guide to running an FBA walks the full process, and the BIP planning canvas is where the pattern becomes prevention, teaching, and response strategies. In Evident, ABC entries are logged against the student next to their daily goal data, so the incidents and the trend come out together when the team builds a packet.
If you would rather print a ready form, the free ABC data sheet has time, setting, antecedent, behavior, consequence, and got or avoided columns. Once the team agrees on the function, the free BIP draft builder turns it and the chosen strategies into a one-page draft for the team to review.
Common questions
How many rows should I fill in before the team looks at the log?
Read it after every session, but plan on at least three to five sessions across different settings before anyone names a function. The stopping point is a pattern that repeats across settings and observers, not a row count.
Should the log include times when the behavior did not happen?
Yes, when the likely trigger was present. A row that says an independent task was assigned and the student stayed in seat is as useful as a row where the behavior happened, because it tests whether the antecedent is really doing the work.
Can a paraprofessional or co-teacher use the same log?
Yes, if everyone uses the same behavior definition from the header. Before relying on a second observer, watch one session together and compare rows. If your entries describe different events, tighten the definition first.