RBT Task List – Measurement (A-03)

We’ve built a solid foundation with A-01: Prepare for Data Collection and A-02: Implement Continuous Measurement Procedures.

Now, we shift our focus to another essential set of tools in the RBT’s measurement toolkit: discontinuous measurement procedures.

Let’s dive into:

A-03: Implementing Discontinuous Measurement Procedures

These methods offer practical ways to gather data when continuous recording of every instance isn’t feasible or is less informative.

Clear Explanation: What is Discontinuous Measurement?

Discontinuous measurement procedures, also commonly referred to as time sampling procedures or interval recording methods, involve observing and recording behavior during specific intervals or at specific moments in time.

Rather than attempting to capture every single instance of the behavior throughout an entire observation period (as is done with continuous measurement).

Because these methods don’t record every occurrence, they provide an estimate of the actual occurrence of the behavior, rather than a complete, exhaustive record.

They are particularly useful and often chosen by supervisors when:

• The target behavior occurs at a very high rate, making it impractical or impossible to accurately count every instance using frequency or rate.
• The behavior is continuous in nature or does not have a clear, discrete onset (beginning) and offset (end) – for example, behaviors like “engagement with a task,” “on-task behavior,” or “stereotypic rocking.”
• The observer (often the RBT) has multiple individuals or multiple behaviors to track simultaneously, making continuous observation of any single behavior unfeasible.
• The primary goal is to get a representative sample of behavior across different time periods or activities.

While discontinuous measurement methods don’t capture every instance, they can be highly efficient and provide valuable data on patterns and trends in behavior when implemented correctly and consistently.

The RBT Task List highlights three main types that RBTs must master: Partial Interval Recording (PIR), Whole Interval Recording (WIR), and Momentary Time Sampling (MTS).

Understanding the distinct rules for each of these procedures is absolutely critical for accurate data collection.

It’s also vital for RBTs to be aware (though the BCBA will handle the interpretation) that these methods often systematically overestimate or underestimate the true occurrence of behavior, depending on the specific method used. This is a key consideration for the BCBA when analyzing the data and making clinical decisions.

Let’s explore each method in detail:

(A-03a) Partial Interval Recording (PIR)

• Detailed Definition:
  In Partial Interval Recording (PIR), an observation period is divided into a series of brief, equal time intervals (e.g., the BCBA might set 10-second intervals for a total observation period of 10 minutes).
  The observer’s task is to record whether the target behavior occurred at any time during each of those short intervals.
  It doesn’t matter how many times the behavior occurred within that interval or for how long it lasted; if the behavior happened even once, for even a fleeting moment, the entire interval is scored as an “occurrence” (often marked with a “+” or a “Yes”).
  If the behavior did not occur at all during the entire interval, it’s scored as a “non-occurrence” (often marked with a “-” or a “No”).
  • Data Typically Reported As: The percentage of intervals in which the behavior occurred. This is calculated by: (Number of intervals behavior occurred / Total number of intervals observed) x 100%.
    • Example: “The target behavior occurred in 7 out of 10 observed intervals, or 70% of intervals.”
  • Tendency of PIR Data: PIR often overestimates the overall duration of a behavior and can also overestimate the frequency of high-rate behaviors. This is because even a very brief occurrence of the behavior results in the entire interval being scored as if the behavior was present throughout.
    If a behavior occurs multiple times in one interval, it’s still just one “+” for that interval.
  • When to Use PIR (Guidance from BCBA):
    • It’s often best for behaviors you want to decrease (e.g., disruptive vocalizations, aggression, off-task behavior).
      The overestimation property can be useful in this context because it makes the data sensitive to even small occurrences of the problem behavior; if the percentage of intervals with the behavior decreases, it’s a good sign.
    • It’s also good for behaviors that occur very briefly or that you might miss if you weren’t watching for that specific interval (as you only need to see it happen once within the interval to score it).
    • It’s useful when it’s difficult to count each instance of a behavior (e.g., high-rate stereotypy) or time its exact duration.
• Real-World Examples of Partial Interval Recording:
  • Tracking Out-of-Seat Behavior in a Classroom:
    • An RBT is observing a student during a 20-minute group lesson. The BCBA has set up the observation period to be divided into 2-minute intervals (resulting in 10 total intervals).
      If the student leaves their designated seat at any point during a 2-minute interval without permission, the RBT marks that interval as an occurrence.
      Even if the student gets up and sits back down 5 times within that single 2-minute interval, it’s still just marked as one “occurrence” for that interval.
    • Result Example: “Out-of-seat behavior occurred in 30% of the 2-minute intervals observed during the lesson.”
  • Monitoring Intermittent Vocal Stereotypy:
    • An RBT is working with a client who engages in brief, intermittent vocal scripting (repeating lines from movies). The session is divided into 30-second intervals for a total of 15 minutes of observation during a play activity.
      If the client engages in any instance of vocal scripting (as per the operational definition) at any time within a 30-second interval, that interval is marked as an occurrence.
    • Result Example: “Vocal scripting was recorded in 12 out of 30 observed intervals (40%).”
  • Observing Aggressive Behaviors in a Group Home Setting:
    • An RBT is observing a resident for aggressive behaviors (defined as hitting or kicking others) during a 1-hour unstructured activity period.
      The RBT uses 5-minute partial intervals. If any instance of aggression occurs within a 5-minute block, that block is scored positively (as an occurrence).
      This helps track whether aggression is present during different parts of the hour without necessarily needing to count every single hit if it’s happening very rapidly in one burst.
• Key Terms/Vocabulary (Specific to PIR):
  • Interval: A short, defined period of time into which the total observation period is divided (e.g., 10 seconds, 1 minute, 5 minutes). All intervals in a PIR system must be of equal length.
  • Scoring Occurrence (for PIR): Marking the interval if the target behavior happened at least once at any point during that interval.
  • Scoring Non-Occurrence (for PIR): Marking the interval if the target behavior did not happen at all during any part of that interval.
  • Overestimation: The inherent tendency of PIR to yield data that suggests the behavior occurred for a greater proportion of time than it actually did, or at a higher frequency than other measures might show.
• Common Mistakes/Misunderstandings (Specific to PIR):
  • Scoring Based on Duration or Frequency Within the Interval: A common error is thinking that if the behavior happened for most of the interval or many times within the interval, it should be scored differently (e.g., with more marks).
    In PIR, any occurrence (even one brief instance) equals a “yes” (or “+”) for that interval.
  • Waiting Until the End of the Interval to Decide/Record: The RBT should record the occurrence on their data sheet as soon as the target behavior is observed within the interval.
    Waiting until the interval is over can lead to forgetting if the behavior actually happened, especially for brief or subtle behaviors.
  • Using Inconsistent Interval Lengths: All intervals within a single observation period must be of equal duration for the data to be meaningful and interpretable.
  • Distraction at Interval Transitions: Missing the precise start or end of an interval due to not having a reliable timing and cueing system.
    A silent, vibrating timer is often best, especially in environments where an audible timer would be disruptive.
  • Using PIR for Behaviors Targeted for Increase: Because PIR tends to overestimate behavior, it’s generally less sensitive for measuring increases in desired behaviors.
    For example, if you are trying to increase “on-task” behavior, PIR might show a high percentage of intervals with on-task behavior even if the student is only briefly on-task in each interval before going off-task again.
    Whole Interval Recording is often better for such targets.

(A-03b) Whole Interval Recording (WIR)

• Detailed Definition:
  In Whole Interval Recording (WIR), the observation period is also divided into a series of brief, equal time intervals (similar to PIR).
  However, the rule for scoring is much stricter: for an interval to be scored as an “occurrence,” the target behavior must occur throughout the entire duration of that interval from the very beginning of the interval to the very end of the interval, without stopping.
  If the behavior stops at any point during the interval, even for a moment, or if it starts after the interval has already begun, that interval is scored as a “non-occurrence” (even if the behavior occurred for 99% of the interval).
  • Data Typically Reported As: The percentage of intervals in which the behavior occurred throughout the entire interval. Calculation: (Number of intervals behavior occurred throughout / Total number of intervals observed) x 100%.
  • Tendency of WIR Data: WIR often underestimates the overall occurrence or total duration of a behavior. This is because if the behavior stops even for a second within the interval, the whole interval is not scored as an occurrence, even if the behavior was present for a large portion of it.
  • When to Use WIR (Guidance from BCBA):
    • It’s often best for behaviors you want to increase, particularly continuous behaviors or states where sustained engagement is important (e.g., on-task behavior, sustained attention, engagement in a cooperative activity, in-seat behavior during a lesson).
      The underestimation property means that if WIR data show an increase in the behavior, it’s a robust and meaningful finding the client is truly sustaining the behavior for longer periods.
    • It’s generally not suitable for behaviors that are very brief or occur sporadically, as they would rarely meet the “throughout the entire interval” criterion.
• Real-World Examples of Whole Interval Recording:
  • Tracking Sustained “On-Task” Behavior During Independent Study:
    • An RBT is working with a student during a 30-minute independent study time. The observation period is divided into 1-minute intervals.
      The RBT scores an interval as an “occurrence” for on-task behavior only if the student remains continuously on-task (e.g., eyes on their work, writing, reading the assigned material, not engaging in off-task conversation) for the entire 60 seconds of that interval.
      If the student looks out the window for 5 seconds during the interval, that interval is marked as a “non-occurrence” for on-task behavior.
    • Result Example: “The student was continuously on-task for the whole interval in 65% of the 1-minute intervals observed.”
  • Monitoring Cooperative Play Between Two Children:
    • An RBT is observing two children who are being encouraged to play cooperatively. The BCBA has set up 30-second whole intervals.
      An interval is scored as a “yes” (occurrence) for cooperative play only if both children are engaged in reciprocal play (e.g., sharing toys appropriately, taking turns in a game, interacting verbally about the play activity) for the entire 30 seconds of that interval.
    • Result Example: “Cooperative play occurred throughout the entire interval in 40% of observed intervals.”
  • Measuring “Remaining in Designated Area” During Circle Time:
    • A client is learning to stay on their designated carpet square during a 10-minute circle time. The RBT uses 2-minute whole intervals.
      An interval is scored as an “occurrence” only if the client remains fully on their carpet square (all body parts within the square) for the entire 2 minutes.
      If they put one foot off the square, even briefly, that interval is scored as a “non-occurrence” for “remaining in area.”
• Key Terms/Vocabulary (Specific to WIR):
  • Throughout the Entire Interval: This is the absolute defining rule for scoring an occurrence in WIR.
  • Continuous Behavior (or State): Behaviors that do not have discrete, easily countable start and end points for each instance, but rather represent ongoing states or activities (e.g., paying attention, sleeping, being engaged).
    WIR is often a good choice for measuring these types of behaviors when sustained engagement is the goal.
  • Underestimation: The inherent tendency of WIR to yield data that suggests the behavior occurred for a smaller proportion of time than it actually did.
• Common Mistakes/Misunderstandings (Specific to WIR):
  • Scoring if Behavior Occurs for Most of the Interval, but Not All: A common error is to confuse WIR with a general impression of the behavior. It must be for the entirety of the interval.
    Even 99% of the interval is not enough to score it as an occurrence.
  • Observer Briefly Looking Away or Stopping Observation Within an Interval: Since the behavior must be continuous throughout the interval for it to be scored, the observer must also be continuously observing the client during that interval.
    If the RBT looks away for a few seconds, they cannot definitively say the behavior occurred throughout the entire interval.
  • Difficulty with Very Short Intervals: If the intervals are extremely short (e.g., 5 seconds), it can be very challenging for an observer to accurately judge if a behavior truly occurred for the entire duration of such a brief period, especially for subtle behaviors.
  • Observer Fatigue: Maintaining constant, unbroken vigilance to ensure the behavior is present throughout every second of many consecutive intervals can be very demanding for the observer.
  • Applying WIR to Low-Rate, Brief Behaviors: If a behavior only happens for a few seconds at a time (e.g., a quick hand raise to ask a question), it will almost never meet the “whole interval” criterion, making WIR an insensitive and inappropriate measure for such behaviors.

(A-03c) Momentary Time Sampling (MTS)

• Detailed Definition:
  In Momentary Time Sampling (also sometimes referred to simply as “time sampling” or “fixed-interval time sampling” in older texts, though MTS is more precise), the observation period is also divided into intervals (often longer intervals than PIR or WIR, e.g., 5, 10, or 15 minutes).
  However, behavior is recorded only at a specific moment, typically at the very end of each interval.The observer looks at the individual at that pre-designated moment (e.g., the exact instant the interval ends, often cued by a timer) and records whether the target behavior is occurring at that precise instant.
  What happens during the rest of the interval (before or after that specific moment) is not considered for scoring that interval.
  • Data Typically Reported As: The percentage of intervals in which the behavior was observed to be occurring at the specific moment of observation. Calculation: (Number of intervals behavior occurred at the moment / Total number of intervals observed) x 100%.
  • Tendency of MTS Data: MTS can either overestimate or underestimate the overall occurrence of behavior, depending on the nature of the behavior and the length of the intervals.
    It is generally considered the least representative or sensitive of the three main discontinuous measurement procedures if one is trying to get a precise estimate of overall behavior occurrence.
    However, it is also often the least demanding and most practical for the observer, especially in busy environments.
  • When to Use MTS (Guidance from BCBA):
    • Excellent for observing multiple individuals simultaneously (e.g., a teacher scanning a classroom and noting how many students are on-task at that moment).
    • Good for observing multiple behaviors of a single individual if the RBT can quickly scan for each at the designated moment.
    • Often used for continuous or state-like behaviors where it’s not critical to know if it happened throughout (e.g., general engagement, on-task behavior, posture).
    • Frequently used when continuous observation is not feasible due to other demands on the observer’s time.
• Real-World Examples of Momentary Time Sampling:
  • Assessing Overall Classroom Engagement (Teacher Application):
    • A teacher wants to get a snapshot of overall student engagement during a lecture. They set a timer for 5-minute intervals.
      At the exact moment the timer sounds (or a silent vibrating alarm goes off), the teacher quickly scans the room and records on a class roster how many students are “on-task” (as per an operational definition:
      e.g., looking at the teacher, taking notes) at that specific instant.
  • Monitoring Stereotypy in a Group Home Setting:
    • An RBT is responsible for several clients in a day program. Every 15 minutes, on the dot (cued by their watch alarm), the RBT briefly observes Client A and notes on their data sheet if Client A is engaging in hand-wringing at that precise moment.
      They then might immediately observe Client B for a different target behavior at that same moment (or a slightly staggered moment), and so on.
  • Measuring In-Seat Behavior During a Mealtime:
    • An RBT is monitoring a child’s in-seat behavior during a 20-minute snack time. The RBT sets a timer for 2-minute intervals.
      At the end of each 2-minute interval (when the timer cues), the RBT looks at the child and records whether the child is seated in their chair at that precise moment.
      Even if the child was out of their seat for 1 minute and 59 seconds of the interval but happened to sit down just as the RBT looked at the designated moment, the interval would be scored as “in-seat” for that moment.
• Key Terms/Vocabulary (Specific to MTS):
  • At the End of the Interval / At the Specific Moment: This is the absolute defining rule for when observation and recording occur in MTS. It’s a snapshot in time.
  • PLACHECK (Planned Activity Check): This is a common variation of MTS used for observing groups of individuals. The observer counts the number of individuals engaged in the target behavior (or the planned activity) at the end of each interval.
  • Observation Moment: The brief, pre-defined instant in time when the observer determines if the behavior is occurring or not.
• Common Mistakes/Misunderstandings (Specific to MTS):
  • Observing for Too Long Around the Designated “Moment”: Instead of taking a quick glance at the designated instant, the RBT might watch for a few seconds before or after the interval officially ends. This effectively turns MTS into a form of short-interval PIR or WIR and defeats the purpose and efficiency of true MTS.
  • Recording Behavior That Ended Just Before or Started Just After the Observation Moment: The rule for MTS is strict: is the behavior happening at that exact instant the observer looks? If it stopped a second before, or starts a second after, it’s a “no” for that moment.
  • Using Predictable Timing Cues for the Client: If the client learns that the RBT only looks or records data when a loud, obvious timer goes off, they might briefly change their behavior around that time (e.g., quickly get on-task just before the beep).
    This is a form of observer/client reactivity that can skew the data. A discrete, unpredictable (to the client) cue for the RBT (like a silent vibrating watch) is often better.
  • Not Recording Immediately at the Observation Moment: The decision about whether the behavior is occurring or not must be made and recorded immediately at that designated moment.
    Delaying the recording can lead to memory errors or confusion.
  • Using MTS for Very Brief, Low-Frequency, or High-Importance Behaviors: Such behaviors are very likely to be missed by MTS because they have to coincide exactly with the observation moment. MTS is generally not suitable for these types of behaviors if an accurate estimate of their occurrence is critical (e.g., you wouldn’t use MTS to measure rare but severe aggression).

General Considerations for All Discontinuous Measurement Procedures

• Observer Training & Interobserver Agreement (IOA): Because these methods require precise timing and strict adherence to specific recording rules, thorough training for the RBT and regular Interobserver Agreement (IOA) checks (where two independent observers collect data simultaneously and compare results) with the supervisor are crucial to ensure that data are being collected accurately and consistently.
• Choosing the Right Interval Length: The length of the observation interval chosen by the BCBA can significantly impact the resulting data. Shorter intervals generally provide more representative and sensitive data but are also more labor-intensive for the observer.
  The supervising BCBA will determine the appropriate interval length based on the nature of the behavior being measured and the practicalities of the observation setting.
• Data Interpretation is for the BCBA: RBTs should understand the inherent biases (i.e., overestimation with PIR, underestimation with WIR, and potential for either with MTS) of each discontinuous measurement method.
  However, the RBT must leave the final interpretation of this data and any clinical decision-making based on it to their supervising BCBA. The RBT’s role is accurate implementation and recording.

Implementing discontinuous measurement procedures, including Partial Interval Recording (PIR), Whole Interval Recording (WIR), and Momentary Time Sampling (MTS), offers practical and efficient ways to gather data when continuous recording isn’t feasible.

These methods are valuable for estimating behavior patterns, especially for high-rate or continuous behaviors, but RBTs must clearly understand the distinct recording rules and inherent biases (overestimation with PIR, underestimation with WIR) of each.

This exploration of A-03 should provide you with a solid understanding of how to apply these sampling techniques correctly. For RBTs, strict adherence to the specific rules for PIR, WIR, and MTS is critical, as these methods provide valuable, albeit estimated, data for the supervising BCBA.

Next up in Section A is A-04: Permanent Product Recording.