Updated for the RBT Test Content Outline (3rd ed., 2026)
RBT data collection turns what you observe into reliable numbers for decision-making. This Unit A study guide covers choosing continuous and discontinuous measurements, permanent products, creating quick summaries, and basic graphs.
RBT Data Collection and Graphing Key Terms
- Continuous measurement: A measure captures every occurrence of a target behavior during observation.
- Discontinuous measurement: A measure samples behavior using time-based procedures.
- Frequency: Counts of a behavior occurs within an observation period.
- Rate: Frequency divided by time.
- Duration: The time of a behavior lasts from onset to offset within a defined observation window.
- Latency: The time between the stimulus and the first instance of the target behavior.
- Interresponse time (IRT): The duration between the end of one response and the beginning of the next.
- Whole-interval recording: Mark an interval as “occurrence” only if the behavior happens for the entire interval.
- Partial-interval recording: Mark an interval as “occurrence” if the behavior happens at any time within the interval.
- Momentary time sampling (MTS): It records whether the behavior is occurring at the exact moment the interval ends.
- Permanent-product recording: The behavior measurements on its lasting outcome (the “product”), which is useful when the result persists and can be counted or checked later.
- Interobserver agreement (IOA) is the degree to which two independent observers’ data match. It is used to evaluate the reliability of measurement.
How RBT Data Collection and Graphing Questions Are Tested on the Exam
RBT Data Collection and Graphing questions usually test whether you can match the right measurement method to the behavior being observed. These items are often less about memorizing definitions and more about recognizing what should be measured and choosing the data system that fits best. In many questions, the behavior is described in a practical way, such as screams 5 times, stays seated for 7 minutes, or starts work 10 seconds after an instruction. That is your clue to identify the correct measurement system. The study guide’s Unit A sections already organize this around continuous measures, discontinuous measures, permanent product, graphing, and data summaries.
A helpful way to think through these questions is to ask: Am I counting how many times the behavior happens, timing how long it lasts, or measuring how long it takes to begin? For example, hits 4 times points to frequency, cried for 3 minutes points to duration, and started cleaning up 6 seconds after the direction points to latency. If the result is something left behind, such as completed homework or a made bed, the best answer is often permanent product recording.
Some questions also test whether you can tell the difference between discontinuous measurement procedures. These items usually give a short scenario and ask how the behavior would be scored under each system. For example, if a student is on task for most of a 1-minute interval but is off task right at the end, partial interval would still be scored yes, whole interval would be no, and momentary time sampling would depend only on whether the behavior is occurring at the exact check moment. The exam is testing whether you know the scoring rule, not just the term.
Graph questions usually test whether you can read simple patterns accurately. You may be asked whether the data show an increasing trend, a decreasing trend, a level change, or high variability after an intervention begins. The best answer is usually a clear description of what the graph shows, without guessing why the change happened. On Unit A questions, the stronger answer stays concrete, matches the behavior to the correct data system, and avoids mixing up similar measurement terms.
Ready to practice? Take the RBT Data Collection and Graphing Quiz (2026). If you’re reviewing older material, you can also try the previous version (2025).
A.1 Continuous Measurement Procedures
Continuous measurement means you record every instance of the behavior during observation. It’s often used when required to capture all occurrences. It gives a complete picture of how often, how long, or how quickly behaviors happen.
- Frequency/Count means how many times a behavior happens.
Example: You tally each time Jordan says “hi” during circle time (7 times).
- Rate equals frequency divided by time.
Example: Emma raises her hand 12 times in 60 minutes. Her hand-raising rate is 12/hour (0.2 hand raises/min)
- Duration tells you how long a behavior lasts.
Example: Maya reads a book for 18 minutes without a break.
- Latency measures the time between an instruction (SD) and the start of the response.
Example: Sam takes 12 seconds to line up after hearing the command “Line up.” The latency is 12 seconds.
- Inter-response Time (IRT) tells you how long it takes for one response to end and the next one to start.
Example: A student finishes one math problem and waits 30 seconds before starting the next one. The inter-response time is 30 seconds.
Common pitfalls
- Forgetting to define the behavior precisely (see A.5).
- Letting conversation or transitions interrupt timing.
Exam Tips
Differences among core continuous methods:
- Frequency focuses on count, so the tool used can be a clicker/tally app.
- Duration, latency, and IRT focus on time, so the tool used can be a timer.
- Rate (count ÷ time) requires both count and time.
Continuous Measurement vs Discontinuous Measurement (see A.2):
- Continuous measurement is designed to capture every moment. If you miss too many instances and capturing them all is not feasible, the measurement would be considered a discontinuous method.
A.2 Discontinuous Measurement Procedures
Discontinuous measurement doesn’t record every occurrence. Instead, it samples behavior. When it is not feasible to observe continuously, this method can be employed (i.e., when being in a busy setting or when a client is exhibiting high‑rate behaviors) or when the sample is sufficient for the measurement goal.
- When using Whole Interval Recording, mark “yes” only if the behavior occurs for the entire interval.
Example: To receive a “yes,” on‑task behavior must continue for the entire 30 seconds during the 30‑second intervals. Score “no” if it lasts less than 30 seconds. This system often underestimates behavior occurrences.
- When using Partial Interval Recording, mark “yes” if the behavior occurs at any time during the interval.
Example: Score “yes” if a client flaps their hand at any point during a 10‑second period. This can overestimate occurrence..
- When applying Momentary Time Sampling (MTS), you’re to check only at the end of the interval.
Example: You look up once every minute to see if the client is on task at that instant. If so, you mark “yes.”.
Common pitfalls
- Changing interval lengths mid‑session.
- Watching continuously during MTS (it’s a spot check, not continuous observation).
Exam Tips
Differences among core discontinuous methods:
- Whole = behaviours occurs the the entire interval;
- Partial = behaviours occurs any time in the interval;
- MTS = present at the moment the interval ends, like taking a photo
Bias & Setups:
- Intervals should be equal and consistent.
- Predefine exactly what counts as behavior before the measurement starts.
A.3 Permanent Product Recording Procedures
Permanent product measures the result of behavior instead of the behavior itself. It is useful when the outcome is measurable and durable.
Examples:
– When taking measurements of routine chores, mark yes if the bed is made or the trash is taken out. No need to record the duration of the chore.
– When taking measurements for homework completion, count the number of correctly completed worksheets. You do not need to observe how the homework was done.
Common pitfalls
- Assuming the learner produced the product without verifying (e.g., a sibling did the chore).
- Ignoring quality/accuracy criteria when the goal specifies correctness, not just completion.
Exam Tips
Be clear when to record:
- Use a permanent product when the result persists after the behavior and can be measured without live observation.
- Define quality criteria (e.g., 9/10 problems correct) if accuracy matters.
- Record opportunities so you can calculate percentages.
Be clear when to record:
- Make sure to record opportunities (this helps to compute a percentage/accuracy if required) and ensure the product could only come from the target behavior.
A.4 Enter Data and Update Graphs
After each session, you would need to enter the data promptly and accurately. Graphs allow the team to “see” behavior change Over time, make data-based decisions.
A line graph is the most typical approach to show behavior data when you only want to make a simple graph. Let’s go over how to read and use the chart below.
This chart records a client’s biting behavior. In this sample, the measurement was taken over one week (Mon-Fri) and captured as five sessions on the X-axis. The behavior was observed and recorded in rate (bites/hour) on the Y-axis. In summary:
X‑axis is used to record time, such as date or session.
Y‑axis is used to record various measures, including frequency, rate, duration, and percentage.
After you plot one point for each session, link them all to make a line. This simple visual helps you consistently track the client’s behavior, especially if there are unusual spikes or dips.
Common pitfalls
- Entering data days later or forgetting to mark the phase change.
- Mixing different measures on the same axis (e.g., some points are rate/hour, others are frequency) without clear labels.
Exam Tips
Understand conditions (phases) on the graph:
- Baseline is data collected before any treatment is introduced. It shows how the behavior naturally occurs without any intervention.
- Intervention is data collected during/after the treatment is introduced. It shows how the behavior looks under the intervention so the BCBA can assess how effective the treatment is.
A.5 Describe Behavior and Environment in Observable and Measurable Terms
When you document behaviors, write down observations and facts, not your assumptions. Use clear and objective language to improve liability.
For instance, don’t say “Sarah was anxious,” states the facts like “Sarah avoided eye contact and spoke softly when answering questions. ” Don’t record “Casey was defiant,” put down what Casey did, such as “When told to ‘clean up,’ Casey walked away and lay on the floor for two minutes.”
Common pitfalls
- Writing interpretations (“didn’t want to listen”) instead of actions.
Exam Tips
Operational definition:
- Always describe topography, context, and clearly defined start/stop criteria in observable terms when recording..
- Ensure to record true behaviors that passed the “dead man’s test.” For example, instead of targeting “being quiet” (which a dead man can do), a better target behavior is “raising a hand before speaking in class.”
A.6 Calculate and Summarize Data
Collected information starts as raw data such as individual counts, time, or yes/no scores. To make it useful and meaningful, you would need to convert the observations into summaries the team can easily interpret. Only when the data has been summarized in a comparable format across sessions does it become valuable for the reports and assessments.
- Rate = Count ÷ Time
Example: if your client has 24 mands in 2 hours, the rate is 12 mands/hour.
- Mean Duration = Total duration ÷ Number of episodes
Example: Your client screamed for 2 minutes, 1 minute, and 3 minutes, for a total of 6 minutes throughout 3 episodes. The average duration equals 2 min/episode.
- Percentage = (Number Correct ÷ Opportunities) × 100
Example: if the count is 16 correct out of 20 trials, the percentage of correctness is 80%
Common pitfalls
- Leaving out context (e.g., reporting a count without a time frame, so rate can’t be properly calculated; reporting numbers correctly without opportunities, so percentage can’t be calculated)
A.7 Trends in Graphed Data
RBTs don’t diagnose or change programs, but you should notice and report patterns to your BCBA. There are a few key visual features to consider when you interpret graphs:
- Trend is the overall direction of data across time. It can be increasing, decreasing, or flat.
- Level represents the overall height of the data path. It can be described as mostly high, mostly low, or moderate.
- Variability describes how spread out the points are.
Now let’s take this knowledge to explain three sample graphs.
- In the graph to the left, the line moves steadily upward across sessions. This indicates an increasing trend.
- In the graph in the middle, the line drops after intervention begins. This indicates a decreasing trend accompanied by a level change at the phase transition.
- In the graph to the right, the points jump around with no clear direction. This pattern is an example of high variability.
Common pitfalls
- Overinterpreting a single unusual data point (an outlier).
Exam Tips
Changes in context:
When interpreting graphs, always check for any context changes (e.g., illness, schedule shifts). If a phase change occurs, describe whether the effect is immediate or gradual when a new condition starts, and explain what you see in clear and plain language.
A.8 Risks of Unreliable Data & Poor Procedural Fidelity
Decisions are only as good as the data they are based on and the integrity with which procedures are carried out. When measurement is unreliable or drifting from the written plan, this can mislead the team, slow down, or even stop progress. This process poses a number of risks, most of which are inaccurate decisions. The data may not accurately reflect what actually occurred, leading to the incorrect modification or continuation of interventions. Decisions based on the incorrect data would not be reliable. Secondly, there may be safety and ethics concerns. Without fidelity, procedures may be ineffective or implemented supportively. Lastly, it’s easy to lose the trust of families and teams as they rely on your accuracy to comprehend the learner’s progress.
To ensure the data quality, always follow the written plan exactly and ask for clarification when anything is unclear. Use operational definitions that are observable and measurable (A.5), select the correct measurement system for the behavior (A.1-A.3), enter data promptly and review for errors (A.4), and participate in treatment integrity checks, and accept feedback with a growth mindset.
Common pitfalls
- Modifying procedures without authorization.
- Choosing a measurement system that doesn’t fit the behavior.
Exam Tips
Understand consequences:
Poor fidelity and unreliable data can lead to misleading decisions, which may also raise ethical concerns and cause delays in progress for clients.
Next Steps
Ready to check your understanding? Take the 2026 RBT Data Collection and Graph Quiz and see where you’re strong and what to review.
Working from older materials? The 2025 (Task List 2.0) version is still available below.