A student displays self-injurious behavior 30 times in an hour. Is that 30 instances spread evenly across 60 minutes, or 30 instances clustered in a 5-minute burst? The frequency is identical, but the pattern—and its clinical implications—are completely different. Inter-Response Time (IRT) captures what frequency misses: the spacing between behaviors.
Definition
Inter-Response Time (IRT) is the elapsed time from the end of one response to the beginning of the next response of the same type.
Why IRT Matters
Consider two scenarios with identical frequency data:
Scenario A: Even Spacing
30 self-injurious hits per hour
Average IRT: 2 minutes
Suggests: Behavior may be maintained by automatic reinforcement (steady pattern)
Scenario B: Burst Pattern
30 self-injurious hits per hour
Average IRT: 10 seconds (during bursts)
Suggests: Behavior occurs in response to specific triggers, then escalates rapidly
The Clinical Insight
These two patterns require different interventions. Scenario A may need sensory-based strategies. Scenario B may need antecedent modifications and de-escalation protocols. Frequency alone cannot distinguish them—IRT can.
When to Use IRT Measurement
IRT is Most Valuable For:
- → Self-injurious behavior: Understanding burst patterns vs. steady rates
- → Stereotypic behavior: Detecting cyclical patterns and triggers
- → High-frequency behaviors: When rate alone misses important patterns
- → DRL interventions: Reinforcing longer time between responses
- → Pacing skills: Teaching appropriate spacing (e.g., requesting breaks)
How to Collect IRT Data
Method 1: Real-Time Recording
- 1. Start a session timer when observation begins
- 2. When behavior occurs, tap/record the timestamp
- 3. Each tap marks the end of one response and enables IRT calculation
- 4. IRT = Time of response N - Time of response (N-1)
Best for: Real-time observation with dedicated observer
Method 2: Video Review
- 1. Record the observation session
- 2. Review video and timestamp each occurrence
- 3. Calculate IRT between consecutive timestamps
- 4. Benefit: Can verify coding and recalculate if needed
Best for: Research, high-stakes decisions, training
IRT Statistics to Report
| Statistic | What It Tells You | Clinical Use |
|---|---|---|
| Mean IRT | Average time between responses | Overall pacing; goal for DRL |
| Median IRT | Middle value (less affected by outliers) | Better "typical" IRT when range is wide |
| Min IRT | Shortest time between responses | Indicates burst severity |
| Max IRT | Longest time between responses | Shows student can refrain for this duration |
| IRT Distribution | Pattern of all IRTs | Reveals bursts vs. steady patterns |
Interpreting IRT Data
Increasing IRT = Progress
When mean IRT grows over time, the student is spacing behaviors further apart.
Example: Baseline mean IRT = 30 seconds
After intervention = 2 minutes
Decreasing IRT = Concern
When mean IRT shrinks, behaviors are clustering closer together (escalation).
Example: Baseline mean IRT = 2 minutes
After stressor = 15 seconds
IRT and DRL Schedules
IRT is the primary measure for Differential Reinforcement of Low Rates (DRL) interventions:
DRL + IRT Example
Target: Reduce frequency of calling out by increasing spacing
Baseline: Mean IRT = 45 seconds (student calls out every 45 seconds on average)
DRL criterion: Reinforcement available if IRT ≥ 2 minutes
Goal: Increase mean IRT to 5 minutes
IRT data directly shows whether the student is meeting the DRL criterion.
Common IRT Patterns
Pattern 1: Tight Clustering (Short IRTs)
Most IRTs are very short (seconds), with occasional long gaps.
Interpretation: Behavior occurs in bursts. May indicate automatic reinforcement or escalation chains.
Pattern 2: Even Distribution
IRTs are relatively consistent throughout observation.
Interpretation: Steady-state behavior. Rate-based interventions may be appropriate.
Pattern 3: Bimodal Distribution
Two distinct clusters: very short IRTs and very long IRTs.
Interpretation: Behavior may be trigger-dependent. Short IRTs during triggering conditions, long IRTs otherwise.
IOA for IRT Data
Use total agreement for IRT statistics:
(Smaller value ÷ Larger value) × 100 = % Agreement
Example: Observer A calculates mean IRT = 45 seconds. Observer B calculates mean IRT = 48 seconds.
(45 ÷ 48) × 100 = 93.75% agreement
Your Next Step
If you have a high-frequency or repetitive behavior that frequency alone hasn't fully explained:
This week: Collect one 10-15 minute observation with timestamps for each occurrence.
Calculate: Mean IRT, min IRT, and max IRT for that session.
Analyze: What does the IRT pattern reveal that frequency did not?
Take Action
Put what you've learned into practice with these resources.
Key Takeaways
- IRT measures the time from the end of one response to the start of the next
- It reveals behavioral patterns that frequency and duration measures cannot detect
- Short IRTs may indicate automatic reinforcement or high-urgency behavior
- Long or increasing IRTs often suggest behavior is coming under better control
- IRT is especially useful for self-injurious, stereotypic, and high-frequency behaviors
Tags:
Ready to Transform Your Classroom?
See how Classroom Pulse can help you streamline behavior data collection and support student outcomes.
Try IRT RecordingFree for up to 3 students • No credit card required
About the Author
The Classroom Pulse Team consists of former Special Education Teachers and BCBAs who are passionate about leveraging technology to reduce teacher burnout and improve student outcomes.
Related Articles
Rate vs. Frequency: When to Use Each Measurement
Understanding the difference between frequency counts and rate calculations is essential for accurate behavior measurement. Learn when raw counts are sufficient and when rate per minute provides more meaningful data for intervention planning.
Multi-Behavior Tracking: Strategies for Complex Cases
When students display multiple target behaviors, tracking them simultaneously requires strategic planning. Learn efficient systems for monitoring several behaviors at once without sacrificing data quality or overwhelming observers.
How to Choose the Right Observation Method: A Decision Guide
With event recording, duration, latency, interval, and permanent product options available, selecting the right measurement method can feel overwhelming. This decision guide walks you through the key questions that lead to the best choice for any target behavior.