Duration in ABA: How to Record, Track, and Calculate It

“Duration” in Applied Behavior Analysis (ABA) therapy refers to how long a behavior lasts. Measuring duration gives clinicians a clearer picture of progress than counting instances would on its own. This guide covers how to record, calculate, and track duration, with practical examples.

What is duration recording in ABA?

Duration recording in ABA measures how long a behavior occurs during a defined period. You start a timer when the behavior begins, stop it when the behavior ends, and record the elapsed time.

Unlike frequency data, duration captures the time a behavior lasts rather than how often it occurs.

For goals like building on-task behavior or reducing the length of outbursts, duration is the metric that tells you whether the treatment approach is working.

When should you use duration recording?

Duration recording is the right choice when time is more clinically relevant than count. Good targets include:

• Time spent engaged in a task or activity
• Length of emotional outbursts
• How long a learner stays seated
• Duration of stereotypy or self-stimulatory behavior
• Time spent in social interaction

These behaviors share two main qualities: a clear start and end, and a length that varies in ways that matter clinically.

When not to use duration recording

Duration measurement isn't a good fit for behaviors that are very brief, occur at a very high rate, or don't have a clean start and end.

Frequency or interval recording are often better ways to measure those behaviors. For example, if a behavior happens very frequently and doesn’t have clear borders, whole-interval, partial-interval, or momentary time sampling tend to be more useful than duration measurement.

How to record duration data in ABA

Here's how to collect duration data, from defining the behavior to calculating your end-of-session metrics.

Step 1: Define the target behavior

Before you start timing, you need a clear, observable definition of the behavior. Everyone on the team, Board Certified Behavior Analysts (BCBAs) and Registered Behavior Technicians (RBTs) alike, should agree on exactly what counts as the start and end point.

For example: "Tantrum is defined as crying, screaming, or dropping to the floor. It begins with the first vocalization or physical drop and ends when the learner has been calm and upright for 5 consecutive seconds."

Vague definitions produce unreliable data. If two observers can't easily agree on when a behavior has started or ended, the definition needs tightening.

Step 2: Choose your recording format

Two of the most common formats for recording duration data are:

• Total duration: The combined time the behavior occurred across all episodes in a session
• Duration per occurrence: The time that each instance or episode lasted, recorded individually

Total duration gives you a session-level summary. Duration per occurrence lets you see patterns within a session, like whether outbursts are getting shorter, even if they're still happening at the same rate.

Step 3: Record each episode

When the behavior begins, start the timer. When it ends, stop it. If the behavior occurs again, restart the timer. Log the start time, end time, and duration of each episode.

Most digital data collection systems come with a simple start/stop button, which cuts the risk of timing errors and automates your calculations.

Step 4: Calculate your summary metrics

At the end of the session, calculate the stats that fit your clinical question best. We’ll cover the key formulas next.

How to calculate duration data

Duration data can be summarized in a few different ways, depending on your clinical question. Here’s how to calculate results using common formulas.

Total duration per session

Add up the duration of each episode during the session.

Example: Three tantrums, lasting 4 minutes, 7 minutes, and 3 minutes, respectively = 14 minutes total duration.

Average duration per occurrence

Divide the total duration by the number of occurrences.

Example: 14 minutes ÷ 3 episodes = 4.67 minutes average per episode.

Percentage of session time

Divide the total duration by the total observation time, then multiply by 100.

Example: 14 minutes ÷ 60-minute session × 100 = 23.3% of session time.

Percentage of session time is especially helpful when observation periods vary in length, since it allows for fair comparisons across sessions.

Duration data examples in ABA practice

Want to better understand how duration data can help with different clinical goals? Here are a few ways you could use this measurement technique.

Example 1: On-task behavior (skill acquisition)

A BCBA wants to build a learner's ability to sit and work independently. At baseline, the learner stays on task for an average of 2 minutes per 30-minute session.

After an intervention, that same learner averages 11 minutes per session. Count-based data wouldn't capture that kind of progress.

Example 2: Outburst reduction (challenging behavior)

A team is tracking how long emotional outbursts last. Frequency data shows the same number of outbursts each week.

But duration data shows that each episode is around 40% shorter than at baseline. Progress like that is easy to miss if you're only counting instances.

Example 3: Social engagement (skill building)

A treatment goal is increased peer interaction, so the team tracks the duration of unprompted social engagement per session.

Over several weeks, they see a steady upward trend in the data. This can then tell the BCBA when to start fading prompts.

Which duration recording format should you use?

Each format has its place, and the right choice depends on your question. Here's a quick guide on when to use each one:

In practice, measuring duration per occurrence is often the starting point: it gives you the data to produce the other formats.

Common duration recording mistakes to avoid

Duration recording is straightforward in theory, but a few common mistakes can compromise your data. Here’s what to keep an eye out for.

Undefined start and end points

If the behavior definition doesn't specify exactly when timing should start and stop, different team members will collect data differently. Any inconsistency means that trends become impossible to see. 

Not resetting the timer between episodes

Some practitioners run the timer continuously across episodes rather than stopping and restarting the clock. This gives you total duration but loses per-occurrence data.

Decide which format you need before the session starts.

Tracking duration for behaviors that don't warrant it

Duration is the wrong measure for very brief, discrete behaviors like hand-raises or verbal mands. Make sure to consider: is time really the most useful dimension to track for this goal?

Collecting data without acting on it

Duration data is only useful if it drives treatment decisions. If you're collecting it every session but not graphing it or adjusting programs based on what you see, it's just paperwork.

Make duration data collection easier with Passage Health

When you're running a program, managing prompts, and keeping a learner engaged, collecting duration data manually leaves a lot of room for error.

Passage Health is an all-in-one platform built to lighten your measurement workload. Here’s what it can do for your team:

• Real-time mobile data collection: On a mobile app, RBTs log duration data that syncs in real time, so nothing gets lost between sessions and notes.
• Automated visualization: Duration trends across sessions are graphed automatically, so BCBAs can easily see shifts in behavior and adjust programs accordingly.
• Flexible programming: BCBAs can build program instructions that are accessible throughout the clinic. Every RBT can work from the same definitions, supporting consistent data collection across staff and settings.
• Customizable session notes: Session note templates are fully configurable, and RBTs can edit notes and collect data at the same time.
• Scheduling: Multiple display options and availability tracking make it easy to schedule the whole team from a single dashboard.
• Integrated billing: Clinical data flows automatically into claims and billing, keeping workflows in one place and removing the need for data re-entry. 
• Reporting and analytics: Track practice performance and utilization across your organization from a single dashboard.
• Frontera AI: Passage Health works alongside Frontera AI, bringing clinical AI features into your ABA workflow, with no need to switch between systems.

New to the platform? Passage Health also offers 1:1 onboarding to get your practice set up quickly.

Book a demo to see how Passage Health can simplify data collection and reduce the admin load for your whole practice.

Frequently asked questions

What is duration in ABA?

Duration in ABA measures how long a behavior lasts, from the moment it starts to the moment it stops. Unlike frequency data, duration data shows time rather than count.

When should you record duration instead of frequency?

Duration recording is the better choice when the length of a behavior is more clinically relevant than how often it occurs. Good candidates include on-task behaviors, social engagement, and emotional outbursts.

How do you calculate average duration per occurrence?

Average duration per occurrence is calculated by dividing the total duration of all episodes in a session by the number of episodes. For example, three episodes lasting 2, 5, and 3 minutes each gives a total of 10 minutes and an average of 3.3 minutes per occurrence.

What's the difference between total duration and duration per occurrence?

Total duration is the combined time a behavior occurred across all episodes in a session, while duration per occurrence records the length of each episode separately. Total duration gives a session-level summary. But collecting per-occurrence data lets you see whether individual episodes are getting shorter or longer over time.

Can duration data be collected digitally?

Yes, duration data can be collected digitally using a mobile platform like Passage Health, where RBTs log data in real time during sessions. Automated graphing makes trends visible across sessions, without any manual calculation.

References

LeBlanc, L. A., Raetz, P. B., Sellers, T. P., et al. (2015). A proposed model for selecting measurement procedures for the assessment and treatment of problem behavior. Behavior Analysis in Practice, 9(1), 77–83. Retrieved from https://pmc.ncbi.nlm.nih.gov/articles/PMC4788644/

Morris, C., Conway, A. A., Becraft, J. L., et al. (2022). Toward an understanding of data collection integrity. Behavior Analysis in Practice, 15, 1361–1372. Retrieved from https://link.springer.com/article/10.1007/s40617-022-00684-x