Partial vs Whole Interval Recording - How to Choose the Right Interval Method

Key Takeaways

Partial interval and whole interval recording are both time-sampling methods, but they serve opposite clinical purposes. Use partial interval recording for behaviors you want to decrease because it overestimates occurrence. Use whole interval recording for behaviors you want to increase because it underestimates occurrence. This conservative approach ensures you can trust your data when making clinical decisions.

The hard part is not just choosing the method. It is running the interval timer while you are still teaching, observing, and responding to the session. TallyFlex supports both Partial Interval and Whole Interval recording with automatic timing, interval placeholders, and real-time percentages, so the method choice can stay tied to the session instead of a separate timer.

If you want the product workflow first, see the Interval Recording feature page. This guide focuses on the method decision.

Interval recording is one of the most practical tools in a behavior analyst’s toolkit. When behaviors occur too rapidly to count or when sustained engagement matters more than frequency, interval recording gives you clinically meaningful data without overwhelming your data collectors.

But choosing the wrong interval type can undermine your entire treatment program. This guide explains when and why to use partial versus whole interval recording, with examples aligned to Cooper, Heron & Heward’s Applied Behavior Analysis (Ch. 4) framework.

Ethics note on decrease-target examples. Per the BACB Ethics Code (2.14, positive-reinforcement-first; 2.15, least-restrictive alternative) and Mathur, Renz & Tarbox (2024), behaviors like stereotypy, scripting, and hand-flapping are not appropriate decrease-targets in marketing-grade examples. The decrease-target examples in this guide use behaviors that meet a harm or learning-access criterion (vocal disruption during instruction, property destruction, high-rate SIB, aggression) and are framed as paired with replacement-behavior teaching (FCT). When you adapt these examples to your own work, the same ethics rules apply.

If you need a refresher on when interval recording makes sense compared to frequency, duration, or latency, see our complete guide to choosing ABA recording methods.

When to Use Interval Recording

Before diving into partial versus whole, let’s establish when interval recording makes sense at all.

How You Arrive at Interval Recording

Interval recording isn’t your first choice - it’s what you use when simpler methods won’t work.

Here’s the decision path that leads you here:

1. You’re tracking behavior occurrence (not skill acquisition like mastery of tasks)
2. You can’t count each instance clearly because the behavior is:
   • Too rapid to count accurately (high-rate vocal disruption during instruction)
   • Continuous with unclear boundaries (sustained engagement, attending)
   • Impossible to distinguish separate occurrences (e.g., bursts of property contact during transitions)
3. Now you need to choose which interval type based on your treatment goal (covered below)

Important: If you CAN count each instance clearly, use Frequency, Duration, or Latency instead. Interval recording is for situations where those methods become impractical.

Use Interval Recording When

• The behavior occurs at such a high rate that counting individual instances becomes impractical (vocal disruption during instruction, high-rate SIB with safety justification)
• You need to measure sustained occurrence over time rather than discrete instances
• Multiple observers need a standardized, efficient method
• You want to capture a representative sample rather than exhaustive data

Consider Alternatives When

• Behaviors have clear starts and ends and occur at countable rates - use frequency
• The exact duration matters clinically - use duration recording
• Response time to instructions matters - use latency recording
• You’re conducting structured teaching trials - use percent correct

Interval recording trades precision for practicality. You sacrifice knowing exactly how many times or how long behavior occurred, but you gain a reliable estimate that multiple observers can collect consistently.

The hardest part of interval recording is managing the timing while also observing and interacting with your student. TallyFlex automates interval timing completely - you choose your interval length, and the app handles the clock so you can focus on observation. Configurable intervals, automatic timestamping, and real-time percentage calculations are all built in.

Partial Interval Recording

How It Works

Divide your observation period into equal intervals. For each interval, mark “yes” if the behavior occurred at any point during that interval, even briefly. Mark “no” only if the behavior never occurred.

Example with 15-second intervals:

You observe a student for 5 minutes (20 intervals) during teacher-led instruction. In interval 3, the student engages in vocal disruption above conversational volume during the teacher’s turn for 2 seconds. You mark that interval as “yes” because the behavior occurred at some point during those 15 seconds.

At the end of your observation, you might have 14 of 20 intervals marked “yes,” giving you 70% of intervals with vocal disruption during instruction.

The Critical Characteristic: Overestimation

Partial interval recording overestimates the actual occurrence of behavior.

Consider: a 1-second occurrence in a 15-second interval scores the same “yes” as 15 full seconds of behavior. If a student vocally disrupted instruction for just 1 second in each of 14 intervals out of 20, partial interval would report 70% - even though the actual behavior occupied less than 5% of the observation period (14 seconds out of 300).

This overestimation is a feature, not a bug, when used correctly.

When to Use Partial Interval

Use partial interval for behaviors you want to decrease (always paired with replacement-behavior teaching - see the ethics note above):

• Vocal disruption during instruction (paired with hand-raising as the replacement)
• Property destruction during transitions (paired with requesting break or movement)
• Aggression during group time (paired with FCT mand for break or attention)
• High-rate self-injurious behavior (SIB) where harm criterion is documented (paired with sensory-replacement and FCT)

Always pair the decrease-target with a named replacement behavior on a parallel tracker. The decrease graph alone misrepresents the intervention; the replacement-behavior graph is what shows the learner gaining a new skill.

Why it works for reduction goals:

When you’re trying to decrease a behavior, you want to be conservative about claiming success. If partial interval shows a decrease from 80% to 40%, you can be confident that actual behavior decreased - probably by more than your data suggests.

This prevents the clinical error of weakening an effective intervention based on artificially inflated improvement.

Whole Interval Recording

How It Works

Divide your observation period into equal intervals. For each interval, mark “yes” only if the behavior occurred continuously throughout the entire interval. Any interruption, even briefly, scores a “no.”

Example with 15-second intervals:

You observe a student for 5 minutes (20 intervals) during independent work, measuring on-task behavior. In interval 3, the student works independently for 14 seconds, then looks around the room for 1 second. You mark that interval as “no” because the behavior did not occur for the entire duration.

At the end of your observation, you might have 12 of 20 intervals marked “yes,” giving you 60% of intervals with sustained on-task behavior.

The Critical Characteristic: Underestimation

Whole interval recording underestimates the actual occurrence of behavior.

Consider: A student on-task for 14.9 of 15 seconds scores “no” for that interval. If the student was genuinely engaged for 95% of the observation but had brief attention lapses, whole interval might report only 60% - significantly lower than reality.

Like partial interval’s overestimation, this underestimation becomes valuable when matched to the right clinical goal.

When to Use Whole Interval

Use whole interval for behaviors you want to increase:

• On-task behavior during instruction
• Sustained engagement with materials
• Appropriate play skills
• Seated behavior during group activities
• Attending to instructor

Why it works for increase goals:

When you’re building a skill or increasing desirable behavior, you want to be conservative about claiming success. If whole interval shows an increase from 30% to 70%, you can be confident the student is genuinely sustaining the behavior longer - likely even more than your data suggests.

This prevents the clinical error of fading support too quickly based on artificially inflated progress.

For implementation details, see our complete whole interval guide.

Decision Framework: Which to Choose

The choice between partial and whole interval comes down to one question: What is your treatment goal?

Behavior Reduction = Partial Interval

Behavior Increase = Whole Interval

The Conservative Principle

Cooper, Heron & Heward (Ch. 4) emphasize that good measurement is conservative about claiming treatment effects. Both interval types achieve this - but only when matched to the correct goal.

The mistake to avoid: Using whole interval for behavior reduction (makes decreases look more dramatic than reality) or partial interval for behavior increase (makes increases look more dramatic than reality).

Practical Implementation Tips

Choosing Interval Length

Start with 15-second intervals for most applications. Adjust based on behavior characteristics:

• Shorter intervals (10-15 seconds): Higher precision, more sensitive to brief behaviors, but more cognitively demanding
• Longer intervals (30-60 seconds): Less demanding on observers, but less sensitive to change

Training Your Team

1. Operationally define the target behavior before starting. Everyone must agree on what counts as “occurring.”
2. Practice with video samples until observers achieve at least 80% inter-observer agreement (IOA).
3. Conduct regular IOA checks during treatment to ensure measurement integrity.

Common Errors to Avoid

• Switching methods mid-program: You cannot compare partial interval data to whole interval data. Choose one and maintain consistency.
• Using the wrong type for your goal: Review the decision framework above before setting up your data system.
• Intervals too long for high-rate behaviors: If behavior occurs multiple times per interval, you lose sensitivity to change.

How TallyFlex Automates Interval Recording

Traditional interval recording requires juggling multiple tasks: watching the student, managing a timer, remembering which interval you’re on, and marking results. This cognitive load can interfere with teaching and interaction.

TallyFlex simplifies this with an inverse approach:

For whole interval: Tap once to start tracking. The app automatically marks each interval as sustained unless you tap to indicate behavior stopped. You focus on the student, not the timer.

For partial interval: Tap when behavior occurs. The app handles interval timing and counting automatically.

Both methods support configurable interval lengths, automatic timestamping, and real-time percentage calculations. Your team collects consistent data while staying engaged in the session.

See our interval recording documentation for implementation details.

Summary

Partial and whole interval recording answer different clinical questions:

• Partial interval tells you whether behavior occurred at all during intervals - use it for behaviors you want to decrease.
• Whole interval tells you whether behavior was sustained throughout intervals - use it for behaviors you want to increase.

Both methods sacrifice precision for practicality, but when matched to the correct treatment goal, they provide the conservative estimates that support sound clinical decisions.

The key is matching the method to the goal: decrease behaviors with partial interval, increase behaviors with whole interval. Get this right, and you can trust your data when it matters most.

What’s Next?

• Compare all nine recording methods: How to Choose the Right ABA Recording Method
• See the TallyFlex interval workflow: Interval Recording feature page
• See whole interval in action: Whole Interval Guide
• Explore all data collection methods: Recording Methods Documentation
• Try interval recording: See how TallyFlex handles interval recording