Partial vs whole interval recording
Partial interval recording marks an interval if the behavior happens at any point in it, so it tends to overestimate. Whole interval recording marks an interval only if the behavior lasts the entire interval, so it tends to underestimate. Both divide the session into equal intervals and report the percentage of intervals scored.
What is partial interval recording?
Partial interval recording splits an observation into equal intervals and scores each one if the behavior occurs at any moment during it. One brief instance is enough. The result is the percentage of intervals scored. Because a split-second behavior fills the whole interval, the data tend to overestimate how much the behavior actually happened.
It is often used for behaviors the team wants to reduce. If the estimate runs high, the team is less likely to think a problem behavior is gone when it is not.
What is whole interval recording?
Whole interval recording scores an interval only if the behavior continues from the first second to the last. A brief break means the interval is not scored. The result is again the percentage of intervals scored. Because short gaps erase whole intervals, the data tend to underestimate how much of the session the behavior filled.
It is often used for behaviors the team wants to increase, such as staying engaged with a task. A low estimate keeps the team from claiming progress too early.
What is the one-question test for partial vs whole interval?
Ask what it takes to score a single interval. If any occurrence at any point is enough, it is partial interval recording. If the behavior must last the full interval without a break, it is whole interval recording. If only the final instant of the interval counts, it is momentary time sampling, a third method.
- Find the rule for marking one interval.
- Any occurrence counts: partial interval, tends to overestimate.
- Whole interval required: whole interval, tends to underestimate.
- Only the last moment counts: momentary time sampling.
Partial and whole interval side by side
| Partial interval | Whole interval | |
|---|---|---|
| Scored when | Behavior occurs at any point | Behavior lasts the entire interval |
| Typical bias | Overestimates | Underestimates |
| Often used for | Behaviors to decrease | Behaviors to increase |
| Example target | Calling out in class | Sustained on-task work |
| Observer attention | Watch the whole interval | Watch the whole interval |
How do you classify interval recording scenarios?
Ignore the behavior named in the scenario and look only at the scoring rule. Both methods can be used with almost any behavior, so the target does not decide the answer. The deciding detail is whether one instance, the full interval or the final moment of the interval earns a mark.
Scenario 1. Intervals are 15 seconds. The technician marks a plus if the learner taps the table even once. Any occurrence counts, so this is partial interval.
Scenario 2. Intervals are 30 seconds. A plus is marked only if the learner keeps writing for all 30 seconds. This is whole interval.
Scenario 3. At the end of each minute, the technician looks up and notes whether the learner is in her seat at that instant. This is momentary time sampling.
Scenario 4. A learner hums for 2 seconds in each of 10 intervals. Partial interval scores all 10, or 100 percent, though humming filled only a small part of the session. This shows the overestimate.
Scenario 5. A learner reads for 28 seconds of every 30-second interval but glances away briefly each time. Whole interval scores none of them, or 0 percent. This shows the underestimate.
How does momentary time sampling compare?
Momentary time sampling scores an interval only if the behavior is happening at the exact moment the interval ends. The observer does not need to watch the whole interval, which makes it practical for groups or busy sessions. It can either overestimate or underestimate, and it can miss brief behaviors that fall between checks.
All three interval methods are discontinuous measures. They estimate behavior rather than counting every instance, which is the job of continuous measures like frequency, rate and duration.
What are common mistakes with interval recording?
The most common mistake is mixing up which method overestimates. Others include reporting interval data as a count of behaviors, scoring an interval twice because the behavior happened twice, changing interval length mid-session, and assuming a high percentage means the behavior lasted most of the session under partial interval recording.
- Remember that each interval gets one mark at most.
- Report a percentage of intervals, not a number of behaviors.
- Keep the interval length the supervisor set.
- Use a timer or vibrating cue so intervals stay accurate.
Practice questions
A learner's humming is recorded with 10-second intervals. The technician marks an interval if humming occurs at any point during it. Which recording method is this, and how does it tend to bias the data?
During a 30-minute session a technician tallies 12 instances of hand-raising. In a 60-minute session the next day she tallies 18. Which measure lets her compare the two sessions fairly?
On a line graph of correct responses per session, the data points rise steadily across the last six sessions after a new teaching procedure started. How is this best described?
Questions people ask
What about momentary time sampling?
It scores only whether the behavior is happening at the instant the interval ends.
Does a shorter interval reduce the bias?
Generally, yes. Shorter intervals tend to make interval estimates closer to the true amount of behavior, at the cost of more observer effort.
Is interval recording continuous measurement?
No. Interval methods sample behavior and give an estimate, so they are discontinuous.
How ready are you?
Ten questions, about 3 minutes, no account. Get your readiness score and your weakest domains.
Check my readiness