> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sightlinebehavior.com/llms.txt
> Use this file to discover all available pages before exploring further.
# ABC Recording
> Document each behavioral event as a three-part sequence (antecedent, behavior, consequence) to build hypotheses about why behavior happens.
* Conducting a functional behavior assessment (FBA)
* Deeper analysis after frequency or interval baseline
* Testing a function hypothesis across sessions
* Informing an intervention plan with ABC contingencies
1. Choose ABC from the method picker
2. Define target behaviors (operational, specific)
3. Optional: pre-load an antecedent / consequence tag set
4. Optional: enable activity context
Instead of tallying or timing, you document each behavioral event as a three-part sequence: what happened before (A), the behavior itself (B), and what happened after (C). Over multiple sessions, these patterns help you develop a hypothesis about why a student displays the behavior.
ABC data alone don't establish causation. They reveal correlational patterns in the natural environment. For maximum defensibility, pair ABC with quantitative baseline data from frequency or interval methods.
## Keyboard shortcuts
| Key | Action |
| --------- | -------------------------------------- |
| `E` | Start a new ABC event |
| `Shift+E` | Duplicate last event's A and C context |
| `1–9` | Toggle tags in current column |
| `Tab` | Advance to next column |
| `Enter` | Commit event (from consequence column) |
| `Cmd+E` | End session |
## Recording ABC events in Sightline
Press **E** to begin entering a new ABC event. A three-column entry form appears: **Antecedent → Behavior → Consequence**. The form defaults to the antecedent column, and the timer keeps running; recording doesn't pause.
The antecedent column shows your pre-loaded tag set (for example "Direct instruction," "Peer interaction," "Transition," "Demand"). Press **1–9** to toggle tags, **Tab** to move to the behavior column (**Shift+Tab** moves back), or type freely to add custom detail. You can select multiple tags for one event.
Press **Tab** to move to the behavior column, which shows your defined target behaviors. Press **1–9** to select one, or type freely. One behavior per event is enforced at the data model level, so the analysis can tie each behavior to its own antecedents and consequences. For co-occurring topographies, define them as separate events.
Press **Tab** again to reach the consequence column, often organized by function (obtain, escape, automatic). Press **1–9** to toggle tags or type freely. You can select multiple consequences if the behavior produced multiple outcomes (for example teacher redirection AND peer attention).
Press **Enter** from the consequence column, or **Cmd+Enter** from any field. The entry commits, the form clears, and you can start the next event. Pressing **Enter** in an earlier column advances the cursor instead of committing. At any point, **?** shows your behavior definitions.
### Duplicating an event's context
If two events share the same antecedent and consequence (same trigger, same outcome, different behavior), press **Shift+E** to duplicate the last committed event's antecedent and consequence context. The behavior field opens blank, so you can quickly select the new behavior and commit.
Example: you recorded "demand → screaming → teacher attention," and immediately the student also hit you during the same demand. Press Shift+E, select "hitting" from the behavior column, and commit. This preserves the fact that both behaviors occurred under identical context, which is useful for contingency analysis.
### Reviewing and editing entries
ABC entries appear chronologically in a timeline below the entry form. Each shows its elapsed time, antecedent tags, behavior name, consequence tags (plus any custom text), and activity context if enabled. Click any entry to edit it: its data loads into the form, and pressing Cmd+Enter saves the changes. Click Cancel to discard the edit.
## Reading the results
An ABC session produces:
* **Event frequency**: total count of each behavior
* **Antecedent patterns**: what triggered each behavior (percentage of instances preceded by each antecedent tag)
* **Consequence patterns**: what followed each behavior (percentage of instances followed by each consequence tag)
* **Function hypothesis**: when the session's antecedent and consequence patterns are strong enough, a synthesized signal that the behavior may be consistent with an escape, attention, tangible, or automatic function. Treat it as a hypothesis to test across repeated sessions, not a conclusion one session can establish; the results screen itself notes that hypotheses confirm across sessions, not within one
These are displayed as a **contingency diagram** — antecedents, behaviors, and consequences connected by weighted paths; hover a tag to isolate its pairings — and a chronological **timeline view** with all details visible. If activity context was enabled, every event carries its activity tag, so you can check whether a contingency holds across settings or is specific to one.
## Common mistakes to avoid
Vague behavior definitions lead to inconsistent data. Spend setup time on clear operational definitions before you start.
Tag sets exist to keep typing minimal during live observation. Tag freely during recording, then add custom notes in analysis if needed.
Record *all* instances, not just the clearest ones. Skipping ambiguous events introduces bias into the antecedent and consequence patterns your analysis rests on.
If two topographies happen together (verbal + physical aggression), record two events under identical antecedent/consequence context using Shift+E. This preserves the analysis granularity you set up for.
## ABC paired with quantitative methods
ABC results pair well with [Frequency](/frequency-recording) or [Interval](/interval-recording) data collected during the same assessment period. Quantitative data shows *how much* behavior occurs; ABC data suggests *why*. Hypothesis-driven intervention needs both. Use frequency or interval data first, then ABC to deepen analysis of the sessions with elevated rates.
***
For turning one session's ABC data into report-ready sentences, including how far a function hypothesis can go in a write-up, see [Writing Up Observation Results](/writing-up-observation-results). For choosing between methods, see [Observation Methods](/observation-methods).