RBT Data Collection & Graphing: What to Know for the Exam

Match the method to the question the data must answer

Data Collection and Graphing is about choosing a measurement that answers the supervisor’s question, then recording it without adding interpretation. Practice Data Collection and Graphing questions in the Passy RBT prep app → before you memorize every abbreviation in a table. Frequency, duration, latency, interval sampling, and permanent product are tools. The item is usually “which tool fits,” not “which term did you highlight.”

The outline assigns Domain A 13 of the 75 scored items (17%). That published weight is not a tally from one person’s form. It is more than Assessment and about the same weight as Ethics. Skipping it because you “already take data at work” is a common miss pattern. Clinic habits can be sloppy relative to the outline.

Continuous, discontinuous, and product

Continuous methods record every instance you can reasonably capture: count (frequency/rate), how long (duration), how long until the behavior starts (latency), or the time between one response and the next (interresponse time). Use them when the behavior is discrete enough to count or time accurately.

Discontinuous methods sample intervals: whole interval, partial interval, momentary time sampling. They estimate. Partial interval tends to overestimate occurrence. Whole interval tends to underestimate. The dedicated comparison is continuous versus discontinuous measurement.

Permanent product records a lasting result (completed worksheet, torn item) when that product is a valid stand-in for the behavior. It is the wrong choice when the product could have been produced some other way.

If two methods both “sort of work,” the stem’s purpose decides. A supervisor who needs an estimate of on-task time is not asking for a frequency tally of every pencil tap.

When interval rules still blur, work a few recording-rule questions in the Passy RBT prep app →.

Graphs, trends, and honest description

Entering data and updating a graph is an RBT task. Reading whether a path is increasing, decreasing, or variable is an RBT task. Inventing a new analysis method is not.

Original sketch, not a BACB item: a client hits a table 6 times during a 30-minute session. The 6 is an observed count. The rate is calculated from that count: 6 ÷ 0.5 hours = 12 responses per hour. The 12 is not a second observation. Across three sessions the calculated rates are 4, 9, then 12 per hour. That path is increasing. One session at 12, after sessions of 11 and 13, is variable, not a new trend. Interresponse time is the time between responses, not the session rate. Interval bias belongs on the continuous-versus-discontinuous page.

Describe behavior in observable terms. “Aggressive” is an interpretation. “Hit the table with an open hand three times” is data. The documentation guide overlaps here because a note that editorializes is both a data problem and a reporting problem.

ABC recording is data collection in the service of later function. The concept page is ABC data and functions. Do not treat ABC as a substitute for the measurement method the program requires.

Watch for fidelity problems the outline cares about: recording late from memory, changing definitions mid-week, or filling intervals you did not observe. Those are how graphs lie without anyone intending fraud.

If a graph trend item feels like guesswork, read explanations in the Passy RBT prep app that name the direction and the variability → instead of staring at a screenshot of someone else’s data sheet.

How Domain A items usually go wrong

They go wrong when the candidate picks the method they use most at work, not the method the stem specified. They go wrong when discontinuous sampling is treated as an exact duration. They go wrong when a label replaces a description.

They also go wrong when Domain A is studied only as vocabulary and never mixed with Acquisition or Reduction. On the real exam, a measurement item can sit next to an ethics item. Mixed practice is the transfer test. The what-to-study guide shows the full six-domain split.

This domain does not require you to design the measurement system. Implement the one in the plan, or ask the supervisor if the plan is silent. That boundary is also an ethics/documentation habit.

When you can name the method and the distortion risk, mix Domain A questions into exam-style practice in the Passy RBT prep app → so graph-reading does not only happen in a measurement chapter.