For Canadian practices outside Quebec
What are the data collection methods in ABA?
The data collection methods in ABA fall into three groups: continuous measures that record every instance (frequency or rate, duration and latency), interval sampling that records behaviour in fixed time windows (partial interval, whole interval and momentary time sampling), and trial-based records such as discrete trials and task analysis. ABC records add the context around a behaviour. The supervising clinician chooses the method that fits the question about each goal.
Last updated: 2026-10-26
By the TherapyCRM team at Data Inn Inc.
Fictional worked example.
Which method answers which question
Define the behaviour and the scoring rule before recording. Each method answers a different question, so results from two methods are not interchangeable.
Swipe or scroll sideways to view all table columns.
| Method | What is recorded | Useful when |
|---|---|---|
| Frequency or rate | Each occurrence; divide by observation time for a rate | The behaviour is brief, with a clear start and end |
| Duration | How long each occurrence lasts | How long matters more than how often, such as time on task |
| Latency | Time from a cue to the start of the response | The concern is how quickly a child starts after an instruction |
| Partial interval | Whether the behaviour occurred at any point in each interval | Counting every instance is impractical; it tends to overestimate |
| Whole interval | Whether the behaviour lasted the entire interval | You want to see a behaviour sustained; it tends to underestimate |
| Momentary time sampling | Whether the behaviour is happening at the end of each interval | The observer is also running the session or watching a group |
| Trial-by-trial | The response and prompt level on each planned opportunity | A skill is taught through structured trials |
| Task analysis | Independence on each step of a sequence | A routine such as handwashing is taught step by step |
| ABC record | Antecedent, behaviour and consequence for each event | The team is exploring what happens around a behaviour |
A fictional example of choosing a method
A team wants to track a fictional child calling out during a 20-minute group activity. A dedicated observer can count every call-out: 8 in 20 minutes is a rate of 0.4 per minute. If the technician is also leading the group, momentary time sampling every 2 minutes is more realistic, and the result is the percentage of 10 checks, not a count. The two figures answer different questions and should not share one graph line.
Where TherapyCRM fits
TherapyCRM Session Runner records trial-by-trial data: each trial as correct, incorrect or no response, with the prompt level, against active targets in the versioned treatment plan. Frequency, duration and interval methods need a separate workflow; the free interval sheet generator, latency recorder and printable ABC, frequency and interval sheets cover them on paper or in the browser.