NDIS Behaviour Data Collection — How to Record Incidents for Your Behaviour Support Plan

Why Most Behaviour Data Is Not Useful

When behaviour data is collected as free text — "participant was unsettled this morning and became aggressive at breakfast" — it cannot be analysed systematically. A BSP can read it, but they cannot count it, graph it, or correlate it with setting events across 50 incidents.

Structured data — where antecedents, strategies, outcomes, and intensity are selected from a consistent list rather than written in prose — can be aggregated, filtered, and graphed. It can answer questions like:

• How many times did this behaviour occur in the last 30 days?
• Is it increasing or decreasing since the support plan changed?
• Which antecedent appears most frequently before this behaviour?
• Which strategies are consistently rated as effective vs ineffective?

These are the questions NDIS plan reviews require evidence for. Prose incident notes cannot answer them.

The ABC Framework — Antecedent, Behaviour, Consequence

ABC recording is the foundation of behaviour data collection in positive behaviour support. Every incident is recorded across three dimensions:

A — Antecedent (what happened before)

The antecedent is the trigger or context immediately before the behaviour occurred. Standard antecedents to capture include:

• Given a task or directive
• Given a correction or limit set
• Preferred activity or item removed or denied
• Transition to new activity or location
• Environmental trigger (noise, crowding, heat)
• Staff change or unfamiliar staff
• Being told No
• Unstructured or waiting time

These should be selected from a consistent list — not written in free text — so frequency analysis is possible.

B — Behaviour (what was observed)

The behaviour description should be in observable, objective terms. "The participant struck staff on the left arm with a closed fist" is recordable and defensible. "The participant was aggressive" is not — it is an interpretation, not an observation.

In addition to the description, capture:

• Behaviour type — physical aggression, verbal aggression, SIB, property damage, elopement, refusal, etc.
• Intensity — a numeric scale (1–5 or 1–6) applied consistently across all incidents
• Duration — using standardised time ranges (<1 min, 1–5 min, 5–10 min, etc.)
• Frequency this shift — once, 2–4 times, 5+ times
• Start and end time — for duration calculation and time-of-day pattern analysis

C — Consequence (what happened after)

The consequence section records the staff response — what strategies were used, how effective they were, and what the outcome was. Capture:

• Strategies used — selected from a standard list (verbal redirection, offered preferred item, adjusted environment, physical intervention, etc.)
• Strategy effectiveness — rated 1–5: 1 = escalated, 3 = some improvement, 5 = fully resolved
• Outcome — redirected successfully, behaviour maintained, behaviour escalated, emergency services called
• Perceived function — attention, escape/avoidance, tangible access, sensory/automatic, communication, physical/medical

Setting Events — The Background Context

Setting events are background factors that increase the likelihood of behaviour but are not the immediate trigger. They need to be captured at the start of every shift, not just when an incident occurs:

• Poor sleep / fatigue
• Physical pain or discomfort
• Missed or late medication
• Feeling anxious or overwhelmed
• Change in routine or staffing
• Sensory overload (noise, crowding)

When setting events are captured systematically, the correlation becomes visible: "This participant's physical aggression incidents cluster on days where poor sleep was recorded the morning before." That is the kind of finding that drives PBS plan changes.

The Problem with Separate Data Collection Systems

Many NDIS providers collect behaviour data in a separate spreadsheet, PDF form, or paper folder — disconnected from the shift record, the clinical notes, and the medication administration record. This creates two problems:

1. Correlation is impossible. You cannot easily compare an incident record with the setting events recorded on the same shift, the sleep data from the previous night, or the medication record from that morning — because they're in different systems.
2. Compliance evidence is hard to produce. When asked for behaviour data for a plan review or audit, you have to manually compile from multiple sources.

When behaviour data is captured in the same system as shift records, notes, and medications, the analysis writes itself.

What BSPs Need From Staff-Recorded Data

Behaviour Support Practitioners use incident data to:

• Demonstrate whether a PBS plan is working (frequency and intensity trending down)
• Identify the function of behaviour for functional behaviour assessment
• Identify which triggers are most common so environmental changes can be prioritised
• Evaluate which strategies are actually effective vs which are in the plan but not working
• Produce evidence for NDIS plan reviews showing progress against goals

All of this requires structured, consistent data. A BSP who receives 50 free-text incident notes must read each one and manually extract the information they need. A BSP who receives 50 structured records can generate a summary report in minutes.

Training Staff to Record Behaviour Data Correctly

The quality of behaviour data depends entirely on the quality of staff recording. Key training points:

• Objective language. Describe what was seen, not what was interpreted. "Struck with closed fist" not "was aggressive."
• Record immediately after the incident. Memory fades quickly and accuracy drops.
• Record even when unsure of the antecedent. "Not recorded" is a legitimate and useful data point — better than guessing.
• Use the standard lists. Consistency matters more than precision. Using the same 15 antecedent options across all staff means the data can be aggregated.
• Record setting events at the start of every shift. Not just when something goes wrong.