Discontinuous measurement procedures involve sampling a target behavior, rather than tracking every instance during the observation period. While these methods are generally considered less precise than continuous measurement, they are highly practical for situations where continuous monitoring isn’t feasible. Each method in discontinuous measurement serves a specific purpose and can still provide valuable insights into behavior trends, making them useful tools for behavior analysts and educators.
In all interval-based data collection procedures, the observation period is divided into specific intervals, such as 10 seconds, 20 seconds, or even 1 minute. The interval duration is typically predetermined by the supervisor or researcher overseeing the process. During each interval, observers record whether the behavior occurred at any point within that time frame, which contrasts with continuous measurement systems where exact frequencies and durations of behaviors are recorded.
While discontinuous systems may not capture every single instance of behavior, they offer a more manageable way to track data when continuous measurement isn’t practical. Below are the common types of discontinuous measurement systems:
1. Partial Interval Recording
Partial interval recording is a method in which the observer notes whether the target behavior occurred at any point during a brief interval of time. If the behavior is observed at any time during the interval, it is marked as a “+.” If the behavior is not observed during the entire interval, it is marked as a “-”.
Strengths:
- Useful for tracking behaviors that are difficult to catch in real-time.
- Allows for data collection without continuous observation.
Limitations:
- Overestimation of behavior: Partial interval recording tends to overestimate behavior since the behavior only needs to occur once during the interval for it to be recorded.
Example: A teacher uses partial interval recording to track a student’s hand flapping behavior. Every time the behavior is observed during the interval, it is marked as a “+.”
2. Whole Interval Recording
Whole interval recording requires the behavior to occur for the entire duration of the interval in order to be counted. If the behavior happens throughout the entire interval, it is marked as a “+.” If the behavior stops at any point, it is marked as a “-”.
Strengths:
- Encourages the observation of sustained behavior.
- Helps track behaviors that should last for a certain duration.
Limitations:
- Underestimation of behavior: Whole interval recording tends to underestimate behavior, as it only counts the interval when the behavior occurs throughout its entire duration.
Example: Whole interval recording can be used to track on-task behavior in a classroom setting. The student is considered “on-task” only if they stay focused for the entire interval, allowing educators to evaluate consistency.
3. Momentary Time Sampling (MTS)
Momentary time sampling records whether the behavior is happening at the end of each interval. The observer checks at the precise moment the interval ends, marking a “+” if the behavior is occurring at that moment or a “-” if it is not.
Strengths:
- Can be used for larger groups or when continuous observation is impractical.
- Simple and quick to implement.
Limitations:
- Provides less detailed data compared to other methods.
- May miss behaviors that don’t align with the specific moment of observation.
Example: Momentary time sampling is used to monitor task engagement in a group of students. At the end of each interval, the observer checks if the students are engaged with the task at hand and marks the appropriate notation.
Conclusion:
Each discontinuous measurement system has its strengths and is suited for different scenarios. Whether it’s partial interval recording for behaviors like vocal stereotypy, whole interval recording for sustained actions like social play, or momentary time sampling for large group activities, these methods offer valuable tools for tracking behavior when continuous measurement isn’t possible. Understanding their limitations—such as over- or underestimating behavior—ensures that the data collected remains as accurate and useful as possible.
