Mastering Discontinuous Measurement: The RBT's Guide to Interval Recording

By

RBT Interval Recording: Decoding Discontinuous Measurement

Data is the absolute bedrock of Applied Behavior Analysis (ABA). Without it, progress is a guess. We have continuous measurement for counting every tiny detail, but discontinuous measurement? That’s about capturing a specific window of time. It’s a core BACB Task List Section A-2 skill. It’s also exactly how we avoid drowning in data while still getting the full picture.


Visualizing the distinct timing of WIR, PIR, and MTS. Alt: Comparative diagram of discontinuous measurement methods for ABA.

A Laptop, a Latte, and a Lab: The Coffee Shop Simulation

Forget the clinic for a second. Imagine you're at a corner table in a coffee shop. You’re practicing. Your "client" is the stranger at the big table. The goal? Measure their "Laptop Use."

Operational Definition: Laptop Use

Definition: Hands—fingers or palms—must be physically touching the keyboard or trackpad. This contact has to be for inputting data or moving things on the screen. Simple as that.

Let's break down Whole Interval Recording (WIR), Partial Interval Recording (PIR), and Momentary Time Sampling (MTS) using this exact scene. It’s easier to learn when it’s real.

1. Whole Interval Recording (WIR)

WIR is brutal. For a behavior to count as a "plus," it has to happen for the entire duration. Every single second. If they stop to blink or scratch an itch, you mark it as a zero. It’s that strict.

The Scenario (WIR)

Set a 30-second timer. Watch for "Laptop Use."

  • Interval 1 (0:00 - 0:30): The stranger types for 25 seconds straight. Suddenly, they take a 5-second sip of their coffee.
  • Result: Non-occurrence (0).

Wait, why a zero? Because 25 seconds isn't 30. WIR requires 100% compliance within that window.

Exam Insight: Because it’s so hard to "win," Whole Interval Recording (WIR) consistently underestimates how much behavior is actually happening. We use this when we want to increase a skill, like "staying on task" or sitting in a chair. If they hit a plus in WIR, they really earned it.

2. Partial Interval Recording (PIR)

PIR is the polar opposite. Did they do it for one second? Did they do it for twenty? It doesn't matter. If the behavior happens at any point, you mark it down. The interval is "done" as soon as the behavior starts.

The Scenario (PIR)

Reset the timer. 30 seconds again.

  • Interval 2 (0:30 - 1:00): The stranger stares at the ceiling for 29 seconds. They look bored. But at the very last second, they tap "Enter."
  • Result: Occurrence (+).

In PIR, that one tap is just as important as a full minute of typing. It’s incredibly sensitive.

Exam Insight: Partial Interval Recording (PIR) overestimates behavior. It makes it look like there’s more going on than there actually is. This makes it perfect for behaviors we want to decrease, like those targeted for extinction (shouting, hitting, etc.).

3. Momentary Time Sampling (MTS)

MTS is the multitasker’s dream. You don't have to watch the client the whole time. You just look up at the exact moment the timer dings. Are they doing the behavior right then? If yes, it’s a plus. If no, it’s a zero.

The Scenario (MTS)

You’re reading your own book now. 30-second timer is ticking.

  • Interval 3 (1:00 - 1:30): They type furiously for 29 seconds. At the 30-second mark, the timer dings. At that exact moment, they pull their hands back to stretch.
  • Result: Non-occurrence (0).

Frustrating, right? They were working the whole time, but they weren't working at the moment you looked. That's the risk of MTS.

Exam Insight: MTS is the most practical choice for busy RBTs or group settings. It’s better for behaviors that persist (like staying on-task) rather than quick, one-off actions. It’s efficient, even if it’s a bit hit-or-miss.

Time to Test Your Logic

Don't let the "over/under" statistics trip you up during the exam. Practice until it's second nature.

Take the Measurement Mock Exam

The "Coffee Shop" Cheat Sheet



Same 30 seconds. Three different results. This table helps you visualize why the method you choose changes the data entirely.

Method Focus Exam "Trap" Best For...
Whole Interval The Entire Time Underestimates behavior Increasing "staying in seat" or "on-task"
Partial Interval Any Part of the Time Overestimates behavior Decreasing problem behaviors like swearing
Momentary Time Sampling The End of the Time Can miss behavior easily Groups or busy environments

Clinical Reality and E-E-A-T

As a BCBA, I might ask you to use discontinuous measurement when we need high-level trends over long periods. Think of token economies: we use WIR to check if a client is earning points for staying on-task for full blocks of time.

But we have to be smart. If hitting is the problem and we use PIR, 10 out of 10 intervals might show a "plus," even if they only hit once every ten minutes. It doesn't track intensity—only occurrence. This is why knowing operational definitions is so vital. It’s about ethical, accurate therapy. For more on how this translates to paperwork, read our guide on writing effective session notes.

Back to Full RBT Study Course

FAQ: Discontinuous Measurement Deep-Dive

What is the main difference between continuous and discontinuous measurement?

Continuous records every single instance (count, duration). Discontinuous only records whether behavior happened within a set window of time, providing a sample instead of a total count.

Why does Whole Interval Recording underestimate behavior?

Because it's "all or nothing." If a behavior happens for 90% of the interval but stops for a second, it's recorded as a zero. The data always looks lower than reality.

When should I use Momentary Time Sampling?

Use it when you can't watch the client 24/7. It's best for long-lasting behaviors (engagement, sitting) and highly practical for group therapy.

Is Partial Interval Recording good for increasing behaviors?

Usually, no. It overestimates. It might make it look like a client is doing a good behavior more often than they really are, which isn't helpful for growth.

What do I need for interval recording as an RBT?

A solid operational definition, a silent/vibrating timer, and a data sheet divided into equal time blocks (10s, 30s, 1m).