Partial interval recording
Partial interval recording is an interval method where you score an interval if the behavior occurs at any moment during it, even for a second. You then report the percentage of intervals scored. It tends to overestimate how much behavior there really was.
What is partial interval recording?
In partial interval recording, you split the observation into equal short intervals, such as 10 seconds each. You mark an interval if the target behavior occurs at least once during it, no matter how briefly. You then divide the scored intervals by the total and report a percentage of intervals.
How do you score partial interval data?
Watch for the whole interval, mark a plus as soon as the behavior appears, and move on to the next. Several occurrences within one interval still count as one. If you observe 40 intervals and score 10, the result is 25 percent of intervals. Interval length stays the same throughout.
Example: with 10-second intervals, a learner spits three separate times within one interval. You score that interval once, not three times, so the data hide the number of events.
Why does partial interval overestimate behavior?
A single brief instance earns the same mark as a behavior that fills the whole interval. That inflates the percentage compared with the true amount of time the behavior takes up. Because of this bias, teams often choose partial interval for behaviors they want to reduce, since an improvement shows up reliably.
- One short hit and a continuous hit both score the interval
- Multiple events in an interval are counted once
- The result estimates, but does not equal, real time or count
When is partial interval recording used?
It is used when counting every instance is not practical, for example with rapid or fleeting behaviors, or when you are also teaching and cannot tally. It suits behaviors the team wants to decrease, because being generous with scoring errs on the safe side. Your supervisor chooses the interval length.
How is it different from whole interval recording?
Partial interval scores the interval if the behavior happens at any point. Whole interval scores it only if the behavior lasts through the entire interval. Partial overestimates the true amount of behavior, while whole underestimates it. Memorize that pair, as the exam likes to test which bias goes with which method.
Exam trap: a question may ask which method suits a behavior to be increased. Whole interval is the cautious choice there, since it demands the behavior last the entire interval.
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?
Using 10-second partial interval recording, a technician sees Lila spit three separate times within the same interval. How should that interval be scored?
A supervisor explains that partial interval recording is often chosen for behaviors the team wants to reduce. What is the main reason?
Using partial interval recording, a technician observes 40 intervals and scores off-task behavior in 14 of them. What percentage of intervals should she report?
Questions people ask
Does partial interval recording give a frequency?
No. It gives the percentage of intervals with the behavior, not a count of events.
Do I choose the interval length?
No. The behavior plan from your supervisor sets it, and you stay with it so data are comparable.
How ready are you?
Ten questions, about 3 minutes, no account. Get your readiness score and your weakest domains.
Check my readiness