ABA glossary
Discontinuous measurement
Also called interval recording, time sampling
Definition
Recording a sample of behavior, usually whether it occurred during intervals or at set moments, instead of every instance. Partial-interval recording, whole-interval recording, and momentary time sampling are the three common types.
In plain words
Recording every instance is sometimes impossible: the behavior is very frequent, lasts a variable time, or the RBT is also teaching. Discontinuous methods divide the session into intervals and record a yes or no for each one. The result is a percentage of intervals, which is an estimate, not an exact count or duration. Each method has a known bias, so the method should match the behavior and the goal. Shorter intervals generally give better estimates but are harder to run.
Example from a session
The BCBA asks the RBT to measure Ava's out-of-seat behavior during a 20-minute class using 1-minute intervals. Instead of timing every episode, the RBT checks at each vibration of her timer and marks whether Ava is out of her seat at that moment. Twenty intervals, 5 scored: 25% of intervals. The RBT reports it as "25% of intervals," not "Ava was out of her seat 25% of the time."
For BCBA candidates and supervisors
Analysts select a discontinuous method by weighing the behavior's dimension, the direction of desired change, and the bias: partial-interval recording overestimates duration (useful for behavior to decrease), whole-interval recording underestimates it (useful for behavior to increase), and momentary time sampling can err either way and misses brief behaviors. Interval length should be short relative to the behavior's typical episodes. Analysts check periodically against a continuous measure when decisions depend on precise values, because discontinuous data can show change that did not happen or hide change that did.
Often confused with continuous measurement, momentary time sampling
Continuous measurement records every instance (count, rate, duration, latency, IRT). Discontinuous measurement records only whether behavior occurred in or at each interval, so it yields an estimate. If the stem says the observer records every occurrence, it is continuous; if it divides the session into intervals, it is discontinuous.
Exam tip
Look for intervals or checks at set times; that is discontinuous. Then decide which type by the scoring rule: "any time during" is partial interval, "the entire interval" is whole interval, "at the moment the interval ends" is momentary time sampling.
Can you spot discontinuous measurement in a scenario?
Practice RBT A.2 with exam-style questions and instant feedback.