Measuring Behavior: Frequency, Duration, Latency and Rate

The question decides the measurement

Two observers watch the same learner hum during a work period. One writes, "Hummed 6 times." The other writes, "Hummed for 4 minutes in total." Both notes are accurate, and both describe the same humming. They answer different questions about it: how many times, and for how long.

By the end of this lesson, you should be able to choose frequency, duration or latency for a clearly defined behavior, calculate rate with its units, and compare sessions of different lengths. You only need counting, division, and converting between seconds, minutes and hours.

Three ways to measure one behavior

Each measure needs a behavior defined clearly enough that an observer can see when it starts and stops.

Frequency means how many times the behavior happened. It fits brief behaviors with a clear beginning and end, such as clapping once or pressing a buzzer.

Duration means how long the behavior lasted, from its start to its end. It fits behaviors that continue over time, such as pacing around the room.

Latency means how long the learner waited after an instruction or cue before starting the behavior. It measures the wait before the behavior, not the behavior itself.

Changing one fact changes the measure. Suppose an adult says, "Put on your shoes." If you time from the instruction until the learner touches a shoe, you are measuring latency. If you time from the first touch until both shoes are on, you are measuring duration. A common mistake is to call the whole stretch "duration." The wait comes first, and that wait is the latency.

Rate: count per unit of time

A count alone does not tell you how often something happens. Ten instances in 10 minutes is very different from ten instances in a full day. Rate fixes this. To find rate, divide the count by the observation time, and keep the units: instances per minute or per hour. The observation time is the period you were actually watching and able to record the behavior.

Worked example 1

During a 45-minute observation, a learner taps the desk 18 times.

Rate = 18 taps ÷ 45 minutes = 0.4 taps per minute.

Taps divided by minutes gives taps per minute. To check, multiply back: 0.4 × 45 = 18.

A tempting mistake is to divide by the time the tapping itself took. If all the taps added up to 3 minutes, then 18 ÷ 3 = 6 is not the rate. Rate uses the observation time, because tapping could have happened at any point during the 45 minutes you were watching.

Worked example 2

On Monday, a learner makes 12 requests in a 20-minute session. On Tuesday, the learner makes 15 requests in a 60-minute session. The raw counts suggest Tuesday had more requesting.

Monday: 12 ÷ 20 = 0.6 requests per minute.

Tuesday: 15 ÷ 60 = 0.25 requests per minute.

Monday's rate is more than twice Tuesday's. The bigger count came from a longer session, not from more frequent requesting. Before you compare, make sure both rates use the same time unit.

Checkpoint

A learner repeatedly asks, "Is it time to go?" In Session A, observed for 2 hours, the learner asks 30 times. In Session B, observed for 40 minutes, the learner asks 12 times. Which session had the higher rate?

Then: an adult says, "Line up." The learner steps toward the door 12 seconds later and takes 20 more seconds to reach the line. Which number is the latency, and which is the duration?

Answer: Session B had the higher rate; 12 seconds is latency and 20 seconds is duration

First convert 2 hours to 120 minutes. Session A: 30 ÷ 120 = 0.25 questions per minute. Session B: 12 ÷ 40 = 0.3 questions per minute. Session B is higher even though its count is smaller.

If you divided 30 by 2, you got 15 per hour. That is correct, but it cannot be compared directly with 0.3 per minute. In per-hour units, Session B is 0.3 × 60 = 18 per hour, which is still higher.

The 12 seconds is latency because it is the wait between the instruction and the first step. The 20 seconds is duration because it is how long lining up took once it began. Adding them into 32 seconds of "duration" would mix the wait with the behavior.

What to remember

Use frequency when you count, duration when you time the behavior from start to end, and latency when you time the wait before it starts. Rate is the count divided by the observation time, written with its units. To compare sessions of different lengths, compare rates in the same unit, not raw counts.

Sources

Behavior Analyst Certification Board. (2023, updated 09/2024). RBT Test Content Outline (3rd ed.). https://www.bacb.com/wp-content/rbt-outline-3rdEd/

IRIS Center, Vanderbilt University. Measuring Behavior: Systematic Direct Observation. https://iris.peabody.vanderbilt.edu/case-study/measuring-behavior/

Independent study resource

This study resource is not affiliated with, sponsored by, or endorsed by the Behavior Analyst Certification Board (BACB). BACB, Registered Behavior Technician and RBT are registered trademarks of the Behavior Analyst Certification Board. Source citations do not imply endorsement by their authors or publishers.