Partial interval, whole interval and momentary time sampling compared
By the IEP Atlas editorial teamLast updated
Short answer
Partial interval recording marks an interval if the behavior happens at any point in it. Whole interval marks it only if the behavior lasts the entire interval. Momentary time sampling marks it if the behavior is happening at the instant the interval ends. Partial tends to overestimate, whole tends to underestimate, and momentary is easiest while teaching.
Interval recording splits an observation into equal chunks of time and scores each chunk yes or no. The result is a percent of intervals, which estimates how much of the period the behavior filled. The three versions differ only in the rule for marking a chunk, but that rule changes the number you get.
Use interval recording when a behavior is too frequent or too continuous to count or time one episode at a time, such as engagement over a 30-minute block. For where it fits among all methods, see the IEP data collection guide and the interval recording measurement page.
The deciding question
Ask: do I want to see this behavior go up or go down, and can I watch the whole time?
- If you want the behavior to decrease and can watch continuously, use partial interval. It tends to overstate the behavior, so a drop is a conservative sign of real change.
- If you want the behavior to increase and can watch continuously, use whole interval. It tends to understate the behavior, so growth is not inflated.
- If you are teaching while you observe, or watching several students, use momentary time sampling. You only need to look at the moment each interval ends.
Side-by-side comparison
| Feature | Partial interval | Whole interval | Momentary time sampling |
|---|---|---|---|
| Scoring rule | Mark + if the behavior occurs at any point in the interval | Mark + only if the behavior lasts the entire interval | Mark + if the behavior is occurring at the instant the interval ends |
| Direction of error | Tends to overestimate | Tends to underestimate | Can err either way; closer with short intervals |
| Basis in law or practice | IEP must state how progress is measured;[1] interval methods are standard in behavior analysis | Same | Same |
| Who decides | IEP team, often on the BCBA's or school psychologist's advice | IEP team, often on the BCBA's or school psychologist's advice | IEP team; a practical choice when the teacher collects |
| Observer attention | Continuous | Continuous | Only at the end of each interval |
| Best for | Behaviors to reduce: calling out, stereotypy, off-task talk | Behaviors to increase: engagement, staying in area, cooperative play | Long periods, several students, teachers who are also teaching |
| Example | Talked to a peer at some point in 14 of 20 intervals = 70% | Engaged for all of 6 of 20 intervals = 30% | Engaged at 11 of 20 interval ends = 55% |
| Data sheet | Interval recording data sheet | Same sheet, whole rule | Same sheet, momentary rule |
| Where documented | Goal criterion, BIP and observation log | Goal criterion, BIP and observation log | Goal criterion and observation log |
The example row comes from a single fictional 20-minute observation of one student, scored three ways. The same behavior produced 70%, 30% and 55%. That spread is why the goal must name the method and why you can never switch methods mid-year without a new baseline.
How each method works
Partial interval recording
Set a timer to the interval length, often 10 to 30 seconds. Watch the student through each interval. If the behavior happens even once, mark + and wait for the next interval. A student who calls out once in a 15-second interval and a student who talks the whole 15 seconds both get +.
That is why partial interval inflates the picture. It is a good fit for behaviors you want to reduce, since showing a decrease under a method that leans high is a safe claim.
Whole interval recording
Watch through each interval. Mark + only if the behavior continued from the start to the end of the interval with no break. A student who is engaged for 14 of 15 seconds gets a minus.
Because one glance away costs the whole interval, whole interval leans low. It suits goals about sustaining a behavior, like staying engaged, because the reported growth is not exaggerated. Keep the intervals short. A 5-minute whole interval is almost always scored minus.
Momentary time sampling
Set a vibrating timer for the interval, often 1 to 5 minutes. When it buzzes, look at the student and mark whether the behavior is happening right then. Between buzzes, keep teaching. The momentary time sampling glossary entry has the short definition.
This is the method most teachers can actually run. It also lets you rotate across students: buzz one, look at Student A; buzz two, look at Student B. Shorter intervals give a closer estimate, but each check takes your attention for a second.
Three classroom scenarios
Scenario 1: off-task talking in a co-taught class
A BCBA is observing a 7th grader who whispers to peers during instruction. The team wants this to go down. The BCBA can watch for the full 30 minutes.
Answer: partial interval with 15-second intervals. The goal is a decrease and the observer can watch continuously. Baseline across three observations: talking in a median of 62% of intervals. The goal targets no more than 20%.
Scenario 2: engagement during independent work
A 3rd grader's goal is to stay engaged during independent math. A para supports three students during the same block and cannot watch one student without a break.
Answer: momentary time sampling at 2-minute intervals across a 20-minute block, giving 10 checks per session. The para rotates looks among the three students on staggered timers. Baseline: engaged at a median of 4 of 10 checks, 40%.
Scenario 3: cooperative play at recess
A kindergartner's goal is to play with a peer for longer stretches. A school psychologist observes for 10 minutes with nothing else to do.
Answer: whole interval with 30-second intervals. The goal is about sustaining the behavior and the observer watches continuously. Baseline: 3 of 20 intervals, 15%. A whole interval number that rises is strong evidence that play really is lasting longer.
Choosing the interval length
Short intervals give more precise estimates and more marks per session. Long intervals are easier to run. A few rules of thumb from practice:
- Partial and whole interval: 10 to 30 seconds, with a dedicated observer.
- Momentary time sampling: 1 to 5 minutes for a teacher who is also teaching.
- Keep it fixed. Use the same interval length and the same observation length every session, or the percents will not compare.
Write both on the interval recording data sheet before the first session, along with the behavior's definition. Leave a box blank if you could not observe, and divide by the intervals you actually scored.
When interval recording is the wrong choice
If the behavior is brief and countable, use frequency or duration instead. Ten hits in a session is clearer than "hitting in 25% of intervals." If the concern is how fast a student starts, use latency. The guide on how to choose a data collection method walks through each option.
Frequently asked questions
Which interval method is most accurate?
None is exact, because all three estimate rather than measure. Momentary time sampling with short intervals tends to come closest to the true share of time. Partial interval leans high and whole interval leans low, which is why each fits a different kind of goal.
Can I switch interval methods during the year?
Only with a new baseline, because the same behavior can score very differently under each method. Changing the measurement in the goal is also an IEP change, which needs the team's agreement.
How long should an interval observation last?
Long enough to give at least 10 intervals, and the same length every session. For a 15-second partial interval that is a few minutes. For 2-minute momentary samples it is about 20 minutes.
Use it in IEP Atlas
- Interval recording measurement guide
- Interval recording data sheet
- Momentary time sampling in the IEP glossary
- Interval recording in the IEP glossary
- On-task behavior IEP goals
Related guides
- IEP data collection: methods, sheets and schedules that hold up
- Frequency vs duration recording: which one fits the behavior
- How to choose a data collection method for an IEP goal
- Prompt hierarchy examples for IEP data
- Latency recording examples for task initiation and transitions
Sources
- 34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
Written and checked by the IEP Atlas editorial team under our editorial policy. This guide describes general practice and federal rules. It is not legal advice; your state may add requirements, and the IEP team makes the decisions for each student.