If you talk to ABA business owners and clinical directors and ask them how they collect data, you will get very different answers! The response is usually based on how they were trained (reinforcement history, etc.). But is there a best way? What if we are looking just at the question of continuous vs. discontinuous data collection?
If you’re having trouble remembering every definition from the Cooper, Heron & Heward book (as most of us do!), here is a quick refresher on the difference between continuous and discontinuous data collection.
“Continuous measurement is a measurement that detects all instances of the target behavior during the observation period.”* Some examples of continuous measurement are frequency count, rate, and duration.
“Discontinuous measurement is any form of measurement in which some instances of the response class(es) of interest may not be detected.”* Some examples of discontinuous measurement are whole interval, partial interval, and momentary time sampling.
Cooper, Heron & Heward, 2020
So, which is better? Each situation is different and one may be a better fit for different behaviors but we often see that interval data is getting the shaft. This could be that some clinicians shy away from interval data because it can be cumbersome to set up and confusing for BTs to use in some electronic data collection systems.
What is Interval Recording?
In ABA, interval recording divides the observation into shorter periods. During these intervals, observers record whether a specific behavior is present or absent. Similar to taking a series of photos, this method tracks changes in behavior over time. It's especially helpful when observing intermittent behaviors or when continuous observation is not possible.
So what are the perks of choosing discontinuous measures (AKA interval data)? Luckily, this question has already been studied! Here are studies and research findings related to interval data:
The Advantages of Interval Data
- More Efficient for Learner’s Acquisition. Cummings and Carr (2013) found that learners reached acquisition criteria more quickly when the data were measured discontinuously, which resulted in less training (by almost an hour average for 6 of the participants). They found that about a third of trials were maintained better after 3 weeks for the group where continuous measurement was used, but that almost two thirds showed equal maintenance.
- Easier and More Preferred by Clinicians. Taubman et. al. (2013) discovered that continuous recording was most accurate while “estimation” recording was most efficient. They also looked at preference of clinicians, most of whom preferred time sampling.
- Just as Accurate and Informative and Continuous Recording (in some instances). However, Ferguson et. al. (2019) evaluated the accuracy of “estimation data collection” and “trial-by-trial data collection” and found estimation accurately matched trial-by-trial data. They also found that both methods were equally efficient in terms of sessions to mastery and number of trials delivered per session.
With all the benefits of interval recording, we believe clinicians deserve a better way to collect interval data! We spent months of time meeting with clinicians, creating and beta testing a tool that we think will make interval data cool again.
Check out Motivity’s new interval tool. For a preview check out this article.
Types of Interval Recording
Now that you know about the perks on interval recording, understanding which method to use is going to help you along the way to best suit your needs. let’s explore more about its major types and their applications in Applied Behavior Analysis:
Whole Interval Recording:
Whole interval recording involves observing whether a target behavior occurs throughout the entire duration of predefined intervals.
Application: when you want to track behaviors that need to be continuously present throughout a specific time frame this method is your go to.
It is particularly useful for monitoring behaviors such as sustained attention, where you want to ensure the behavior persists for the entire duration of the observation period.
Example: Let's say an ABA therapist is working with an autistic child to improve their attention span during academic tasks. The therapist decides to use whole interval recording to monitor the child's on-task behavior during a 15-minute math lesson.
The therapist divides the lesson into three 5-minute intervals and observes whether the child remains focused on the assigned math problems throughout each interval.
By recording instances of sustained attention, the therapist can assess the effectiveness of interventions aimed at increasing the child's attention span during academic tasks.
Partial Interval Recording:
Partial interval recording is about observing whether a target behavior occurs at any point within predefined intervals, regardless of its duration or frequency.
Application: Employ partial interval recording when you want to track behaviors that may occur intermittently within a given time frame.
This method is beneficial for behaviors that fluctuate in frequency or intensity, allowing you to capture instances of the behavior without requiring continuous observation.
Example: Consider an ABA specialist working with a teenager with oppositional defiant disorder (ODD) to reduce instances of disruptive behavior during a group therapy session.
The specialist opts for partial interval recording to track occurrences of disruptive behavior throughout a 30-minute therapy session. The specialist divides the session into six 5-minute intervals and records whether the teenager engages in disruptive behavior at any point within each interval.
Momentary Time Sampling:
Momentary time sampling is used to record whether a target behavior is present or absent at specific moments within predefined intervals.
Application: Momentary time sampling can help you capture whether a behavior is present at specific moments within an observation period.
This method is suitable for behaviors that occurs poradically or for brief durations, which gives you the ability to obtain a snapshot of behavior patterns without continuous monitoring.
Example: Imagine an ABA practitioner conducting a functional behavior assessment (FBA) for a child with communication difficulties. The practitioner aims to assess the frequency of vocal stereotypy behavior during a 20-minute play session.
Using momentary time sampling, the practitioner sets a timer to prompt observation every 2 minutes and records whether the child exhibits vocal stereotypy behavior at each interval.
Cummings, A. R., Carr, J. E. (2013). Evaluating progress in behavioral programs for children with autism spectrum disorders via continuous and discontinuous measurement. Journal of Applied Behavior Analysis, 42(1), 57–71.
Taubman, M. T., Leaf, R. B., McEachin, J. J., Papovich, S., Leaf, J. B. (2013). A comparison of data collection techniques used with discrete trial training. Research in Autism Spectrum Disorders, 7(9), 1026-1034.
Ferguson, J. L., Milne, C. M., Cihon, J. H., Dotson, A., Leaf, J. B., McEachin, J. Leaf, R. (2019). An evaluation of estimation data collection to trial-by trial data collection during discrete trial teaching. Behavioral Interventions, 35(1), 178-191.
*Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied Behavior Analysis. Pearson.
