What is Partial Interval Recording?
Partial Interval Recording is a discontinuous measurement procedure where the observer marks an interval as "positive" if the behavior occurs at any time during the interval. It does not matter if the behavior happened for 1 second or the entire 10 seconds; if it happened at all, you mark it. For the RBT practice test, remember that this method is best for behaviors you want to decrease.
Because it counts even a fleeting instance as a full "Yes," partial interval recording tends to overestimate the actual occurrence of behavior. This is surprisingly useful clinically. If we overestimate a bad behavior (like hitting) and still see the data go down, we can be very confident the intervention is working.
On the RBT Practice Exam, you will often face scenarios asking which measurement to use for "high frequency" behaviors or behaviors without a clear start and stop. Partial interval is usually the correct answer for these cases.
How to Implement Partial Interval Recording
First, divide your observation period into equal blocks of time (e.g., 10-second intervals). You need a timer that alerts you when each interval ends. During the interval, watch the client.
If the behavior occurs, mark the data sheet immediately and you can stop watching for that specific behavior until the next interval starts. This efficiency is why it is called "discontinuous" measurement—you don't technically have to observe 100% of the time, just until the behavior happens in that interval.
Clinical Use Cases
Partial interval is the gold standard for behaviors like vocal stereotypy (humming, scripting). It is impossible to count every single syllable a child says. Instead, we use 10-second intervals. Did they hum in this 10 seconds? Yes. Did they hum in the next? No.
It is also excellent for disruptive behaviors in a classroom. A teacher cannot count every time a student taps their pencil. But she can look up and see if they are tapping during a checkout interval. It provides a manageable way to track improvement.
The "Overestimation" Rule
You must memorize this for your exam: Partial Interval Overestimates. If a child screams for 1 second out of a 60-second interval, the data shows 100% occurrence for that minute. This makes the behavior look worse than it is (in terms of duration), but it captures the presence of the problem effectively.
Conversely, Whole Interval Underestimates. Keeping these two rules straight is essential for passing your board exam.
RBT Practice Exam Tips for Partial Interval
Look for keywords like "at any point," "at any time," or "overestimates." These are dead giveaways for Partial Interval questions. If the scenario involves a behavior the BCBA wants to reduce, lean towards Partial Interval.
Be careful with "Momentary Time Sampling." Momentary is only at the end of the interval. Partial is any time during the interval. The exam loves to trick you with this subtle difference.
Professional Consistency
When collecting partial interval data, your honesty is paramount. Do not mark "No" just because the behavior was "minor" or "short." If it met the definition, it is a "Yes." Consistency allows the BCBA to see the true trend.
Reporting this data usually involves saying, "Behavior occurred in 30% of intervals." This is easily graphed and understood by parents and insurance providers.
Mastering this skill makes you a valuable asset in complex cases where frequency counts are impossible. It shows you can handle advanced data collection methods with ease.
Frequently Asked Questions About Partial Interval Recording
What does partial interval recording measure?
It measures the presence or absence of a behavior within a specific time block. It marks "Yes" if the behavior happens at all, regardless of for how long.
Does it overestimate or underestimate?
It overestimates the proportion of time the behavior occurs. This is because a 1-second behavior counts as a "full" interval of occurrence.
When should I use it?
Use it for behaviors you want to decrease (like aggression or stereotypy) and for behaviors that happen at such a high rate that counting them (frequency) is impossible.
What is the difference from Whole Interval?
Whole Interval requires the behavior to happen for the entire time to count as "Yes." It underestimates. Partial requires it to happen at any time. It overestimates.
How do I calculate the percentage?
Divide the number of "Yes" intervals by the total number of intervals. 5 "Yes" intervals out of 20 total = 25% occurrence.
What tools do I need?
An interval timer (app or device) is critical. You cannot estimate 10 seconds accurately in your head while watching a client. You also need a data sheet or tablet.
Is it continuous measurement?
No. It is discontinuous measurement. You are breaking the session into chunks (intervals) rather than recording every single instance in real-time continuous flow.
Can I listen instead of watch?
For vocal behaviors (like screaming), yes. But generally, you need to observe. If you look away and the behavior happens, you missed the data point.
What if the behavior starts in one interval and ends in the next?
Then you mark both intervals as "Yes." If it happened in Interval A, A is Yes. If it continued into Interval B, B is also Yes.
Why is it good for "decrease" goals?
Because it is a "strict" measure. To get a "No," the client must be good for the entire interval. It sets a high bar for behavior reduction.
What logic should I use on the RBT exam?
If the behavior is "bad" and high-frequency, pick Partial Interval. If the behavior is "good" and you want to build endurance, pick Whole Interval.
Is Momentary Time Sampling easier?
Yes, because you only look at the end. But Partial Interval is more accurate than Momentary for most high-rate disruptive behaviors.