22 Measurement & Data Collection Practice Questions & Answers

Every Measurement & Data Collection practice question from the RBT Practice Test, with the correct answer and a short explanation.

Start practice test →

1. An RBT gives the instruction "Put on your shoes." The client eventually complies, but the supervisor is concerned about how long the client takes to begin responding after an instruction is delivered. Which measure BEST captures that concern?
- A.The duration of the response
- B.Inter-response time
- C.Latency✓ Answer
- D.The frequency of responses
Latency is the time that elapses between the antecedent event (here, the instruction) and the beginning of the response, so it is the measure that answers "how long before the client starts?" Duration times one instance from onset to offset, and inter-response time measures from the end of one response to the start of the next, so neither one isolates the delay after an instruction.
Source: BACB RBT Task List, Measurement — continuous measurement procedures (frequency, duration, latency, inter-response time)Report a problem with this question
2. A client takes bites of food one right after another with almost no pause between them. The supervisor wants a measure of how much time passes between the end of one bite and the beginning of the next. Which measurement procedure should the RBT use?
- A.Latency
- B.Whole-interval recording
- C.The duration of each bite
- D.Inter-response time✓ Answer
Inter-response time is defined as the time from the end of one response to the onset of the next response, which is exactly the pacing question the supervisor asked. Latency is measured from an antecedent to a response, and duration is measured from the onset to the offset of a single response, so neither describes the spacing between successive responses.
Source: BACB RBT Task List, Measurement — continuous measurement procedures (inter-response time)Report a problem with this question
3. A client's tantrums have not changed in number from week to week, but the RBT notices that some episodes last only seconds while others go on much longer, and the supervisor wants to know whether episodes are becoming shorter. Which measure BEST answers that question?
- A.Permanent product recording
- B.Duration✓ Answer
- C.Latency
- D.Frequency
Duration measures the time from the onset to the offset of each episode, so it is the only one of these measures that can detect episodes getting shorter while their number stays the same. Frequency counts episodes and would show no change at all, which is why duration is chosen for behaviour that varies in length.
Source: BACB RBT Task List, Measurement — selecting duration for behaviour that varies in lengthReport a problem with this question
4. An RBT writes "hand raising: 12" on the data sheet but does not record how long the observation lasted. Why is this a problem?
- A.A count may only be reported as a percentage of intervals
- B.Hand raising must always be measured with whole-interval recording
- C.Without the length of the observation period the count cannot be converted to a rate, so it cannot be meaningfully compared with other sessions✓ Answer
- D.Counting samples behaviour rather than recording it, so counts are never reported alone
A raw count only becomes interpretable when it is paired with the time over which it was collected, because rate is count divided by observation time. Twelve responses in a short session and twelve in a long session describe very different behaviour, so without the observation time the supervisor cannot compare sessions.
Source: BACB RBT Task List, Measurement — count and rate (count must be paired with observation time)Report a problem with this question
5. A supervisor asks an RBT to record two things: the time from the moment a timer sounds until the client begins washing his hands, and the time from the end of one instance of hand flapping until the beginning of the next instance. Which two continuous measures has the supervisor requested, in that order?
- A.Latency, then inter-response time✓ Answer
- B.Inter-response time, then duration
- C.Duration, then latency
- D.Latency, then duration
The first request runs from an antecedent event to the onset of a response, which is latency; the second runs from the offset of one response to the onset of the next, which is inter-response time. Duration would run from the onset to the offset of a single instance and was not requested at all.
Source: BACB RBT Task List, Measurement — distinguishing latency, inter-response time, and durationReport a problem with this question
6. An RBT is told to divide the session into equal intervals and to score an interval as an occurrence if the behaviour happens at any point during that interval. Compared with recording the behaviour continuously, what will these data MOST likely do?
- A.Match a continuous record exactly
- B.Tend to overestimate how much of the session the behaviour actually occupied✓ Answer
- C.Bear no relationship at all to how often the behaviour occurred
- D.Tend to underestimate how much of the session the behaviour actually occupied
This is partial-interval recording, in which a single brief occurrence causes the entire interval to be scored, so the summary credits the behaviour with time it did not actually occupy. That built-in overestimate is why partial-interval data are considered a conservative choice for behaviour targeted for reduction.
Source: BACB RBT Task List, Measurement — partial-interval recording and its direction of errorReport a problem with this question
7. A supervisor wants on-task behaviour measured so that an interval is scored as an occurrence only if the client is on task for the entire interval. What should the RBT expect from the resulting data?
- A.They will tend to underestimate time on task✓ Answer
- B.They will overestimate time on task
- C.They will be unaffected by how long each interval is
- D.They will exactly match a continuous duration record
This is whole-interval recording: the behaviour must last from the start of the interval to its end, so any brief pause makes the whole interval score as non-occurrence and real engagement is under-reported. That systematic underestimate is why whole-interval recording is usually paired with behaviour the team is trying to increase and sustain.
Source: BACB RBT Task List, Measurement — whole-interval recording and its direction of errorReport a problem with this question
8. An RBT is running a group activity and can only glance up briefly at set times. The supervisor asks her to look at the client at the end of each interval and record whether the behaviour is happening at that instant. Which procedure is being described?
- A.Partial-interval recording
- B.Permanent product recording
- C.Whole-interval recording
- D.Momentary time sampling✓ Answer
Momentary time sampling scores only whether the behaviour is occurring at the single moment the interval ends, which is what allows the RBT to keep teaching between observations. Partial- and whole-interval recording both require attention to what happened throughout the interval, not at one instant.
Source: BACB RBT Task List, Measurement — momentary time samplingReport a problem with this question
9. A supervisor wants an estimate of how much of a session a client spends engaged in an ongoing, continuous behaviour, but the RBT must keep teaching and cannot record every instance. Which sampling method is LEAST likely to systematically overestimate or underestimate that behaviour?
- A.Partial-interval recording
- B.Whole-interval recording
- C.Latency recording
- D.Momentary time sampling✓ Answer
Momentary time sampling neither credits a whole interval for one brief occurrence, as partial-interval does, nor discards an interval because of one brief pause, as whole-interval does, so for ongoing behaviour it carries the least systematic bias of the interval methods. Its accuracy still declines as the intervals get longer, because more of the session goes unobserved.
Source: BACB RBT Task List, Measurement — comparing discontinuous measurement proceduresReport a problem with this question
10. Which target behaviour is BEST suited to permanent product recording?
- A.Vocal stereotypy during free play
- B.Remaining attentive during a group lesson
- C.The number of math problems the client completes on a worksheet✓ Answer
- D.Attempts to leave the classroom
Permanent product recording measures the lasting tangible outcome a behaviour leaves behind, so it works only when the behaviour produces something that can be counted afterwards, such as completed worksheet items. Vocal stereotypy, attending, and elopement attempts leave no product to inspect, so they must be observed as they occur.
Source: BACB RBT Task List, Measurement — permanent product recordingReport a problem with this question
11. An RBT measures bed making by inspecting the bed after the client leaves the room. The parent mentions that she sometimes finishes the bed herself when the client is slow. What is the MAIN measurement concern?
- A.Permanent product recording requires the RBT to observe the behaviour as it happens
- B.Permanent product recording requires the RBT to time each interval
- C.Permanent product recording may only be used for academic tasks
- D.The product may not have been produced by the client's own behaviour, so the data may not reflect what the client actually did✓ Answer
Permanent product measurement is valid only when the product was produced by the client's own target behaviour; if another person can contribute to or complete the product, the record no longer measures the client's responding. Not needing to be present is the method's advantage, but that same feature is what makes contamination of the product possible.
Source: BACB RBT Task List, Measurement — validity requirements for permanent product recordingReport a problem with this question
12. An RBT graphs three baseline sessions followed by five intervention sessions on the same line graph. How should the graph be constructed at the point where the intervention began?
- A.Connect the last baseline point to the first intervention point so that the data path stays continuous
- B.Draw a vertical phase-change line between the conditions and do not connect data points across it✓ Answer
- C.Plot the session numbers on the vertical axis so that the change is easier to see
- D.Start an entirely new graph for the intervention data
Data paths are broken at a phase-change line because connecting them would imply that the points were collected under one unchanged set of conditions. Keeping both conditions on the same graph with the path broken lets the supervisor compare level, trend, and variability on either side of the change.
Source: BACB RBT Task List, Measurement — entering data and standard line-graph conventionsReport a problem with this question
13. On a behaviour-analytic line graph, what does a vertical phase-change line tell the reader?
- A.That the data on either side were collected by different observers
- B.That the client met the mastery criterion at that point
- C.That the behaviour changed at that point
- D.That the condition in effect changed at that point✓ Answer
A phase-change line marks only that the conditions in effect changed, for example from baseline to intervention. Whether the behaviour itself changed is a separate question answered by inspecting the level, trend, and variability of the data on each side of the line.
Source: BACB RBT Task List, Measurement — interpreting phase-change lines on graphsReport a problem with this question
14. Which of the following target definitions is written in observable and measurable terms?
- A.The client drops to the floor and screams louder than conversational volume✓ Answer
- B.The client is defiant toward adult instructions
- C.The client seeks attention during group activities
- D.The client becomes frustrated when demands are placed
An operational definition must be objective and complete enough that two people reading it independently would count the same events. "Frustrated", "seeks attention", and "defiant" are inferences about internal states or motives that cannot be seen or heard, whereas dropping to the floor and screaming above conversational volume describes exactly what an observer would see and hear.
Source: BACB RBT Task List, Measurement — describing behaviour in observable and measurable termsReport a problem with this question
15. An RBT is asked to review four proposed target behaviours. Which one fails the Dead Man's Test and therefore is not written as a behaviour?
- A.Raising a hand
- B.Not talking during silent reading✓ Answer
- C.Walking to the sink
- D.Handing a picture card to an adult
The Dead Man's Test says that if a dead person can do it, it is not behaviour. A dead person can perfectly satisfy "not talking", so that target should be rewritten to state what the client actively does instead, such as reading with eyes on the page or raising a hand to comment.
Source: BACB RBT Task List, Measurement — Dead Man's Test for target behaviour definitionsReport a problem with this question
16. An RBT records 20 occurrences of hand raising during a 40-minute session. The supervisor asks for the rate per hour. What should the RBT report?
- A.0.5 per hour
- B.20 per hour
- C.30 per hour✓ Answer
- D.40 per hour
Rate is the count divided by observation time, expressed in the unit the supervisor requested: 20 responses in 40 minutes is 0.5 per minute, and 0.5 multiplied by 60 gives 30 per hour. Reporting the raw count of 20 skips the unit conversion, and 0.5 is the per-minute figure rather than the per-hour figure.
Source: BACB RBT Task List, Measurement — calculating rate (count divided by observation time, with unit conversion)Report a problem with this question
17. During one session a client had 4 tantrum episodes and the recorded durations totalled 12 minutes. What is the mean duration per episode?
- A.12 minutes
- B.48 minutes
- C.4 minutes
- D.3 minutes✓ Answer
Mean duration is total duration divided by the number of occurrences, so 12 minutes divided by 4 episodes is 3 minutes per episode. Total duration alone describes the whole session rather than the typical episode, which is why the two summaries answer different clinical questions.
Source: BACB RBT Task List, Measurement — summarizing data as mean durationReport a problem with this question
18. Aggression is targeted for reduction. Over the last ten sessions the graphed data have moved steadily downward, with the points staying close to that downward path from session to session. How should the RBT describe this data path when reporting to the supervisor?
- A.An increasing trend with low variability
- B.A decreasing trend with low variability✓ Answer
- C.A change in level with no trend
- D.A flat trend with high variability
Trend describes the direction of the data path and variability describes how much the points bounce around it, so a steady downward path with tightly clustered points is a decreasing trend with low variability. Direction must always be read against the goal: decreasing is improvement for a reduction target but would be a problem for a skill being built up.
Source: BACB RBT Task List, Measurement — identifying trend, level, and variability in graphed dataReport a problem with this question
19. The percentage of correct responses on a skill the client is learning has declined across six consecutive sessions. What is the MOST appropriate action for the RBT?
- A.Explain to the family what the decline means for the client's progress
- B.Continue collecting and graphing the data accurately and report the declining trend to the supervisor promptly✓ Answer
- C.Change the prompting procedure to reverse the decline
- D.Stop graphing the programme until the data improve
Noticing and describing a trend is part of the RBT's data-collection role, but deciding what it means and what to do about it is the supervising behaviour analyst's decision. For an acquisition target a declining trend signals that the programme may need review, so the RBT keeps the record accurate and notifies the supervisor promptly rather than modifying the procedure or interpreting the data for the family.
Source: BACB RBT Task List, Measurement — reporting trends to the supervisor; RBT scope of practiceReport a problem with this question
20. An RBT is very busy during sessions, so instead of recording as the behaviour occurs she writes down her best recollection of the counts at the end of the day. What is the PRIMARY risk of this practice?
- A.Data reconstructed from memory may not accurately reflect what occurred, and the supervisor may then make clinical decisions based on inaccurate data✓ Answer
- B.There is no real risk as long as the totals are close to what happened
- C.It only becomes a problem when duration rather than count is being measured
- D.It only matters on days when a second observer is also collecting data
Data are only useful to the degree they are an accurate record of what the client actually did, and memory at the end of a day systematically loses and distorts occurrences. Because the supervising behaviour analyst uses those numbers to judge whether a programme is working, unreliable data can keep an ineffective programme running or lead to stopping one that was working.
Source: BACB RBT Task List, Measurement — risks associated with unreliable data collectionReport a problem with this question
21. A fire drill interrupts a session and no data were collected for one of the client's programmes. What should the RBT do?
- A.Carry the previous session's data forward so the graph has no gap
- B.Leave the data cell blank, note why data were not collected, and inform the supervisor✓ Answer
- C.Estimate a value based on what the client usually does
- D.Record zero occurrences for that programme
No observation took place, so no value was obtained; recording a zero would falsely assert that the behaviour was observed and did not occur, and estimating or copying forward invents data that were never collected. Leaving the cell blank with a note and telling the supervisor keeps the record truthful and lets the supervisor account for the missing session.
Source: BACB RBT Task List, Measurement — accurate data recording; never estimating or fabricating missing dataReport a problem with this question
22. A material required by a written teaching protocol is missing, so an RBT delivers the procedure a different way for the whole session and does not mention it to anyone. What is the MOST significant risk of doing this?
- A.The procedure was not delivered as written, so the data no longer reflect the programme as designed and the supervisor may draw the wrong conclusion about whether it works✓ Answer
- B.The session's data automatically become a permanent product measure
- C.Procedural fidelity matters only during assessment, so nothing is affected
- D.The RBT must now redraw the graph with a new phase-change line
Procedural fidelity means the protocol is delivered exactly as written; when implementation drifts, the resulting data describe some other procedure while still being graphed as if they described the programme. The supervisor then judges effectiveness from a mismatch, so an RBT who cannot run a procedure as written records what happened objectively and tells the supervisor promptly rather than improvising silently.
Source: BACB RBT Task List, Measurement — risks associated with poor procedural fidelity; RBT scope of practiceReport a problem with this question

Practice questions based on the BACB RBT Task List and standard applied behavior analysis practice. RBT and BACB are marks of the Behavior Analyst Certification Board; this site is not affiliated with or endorsed by the BACB. Requirements, training rules, and the task list itself are revised periodically — confirm current requirements with the BACB and your supervisor before testing. About the RBT credential →

1. An RBT gives the instruction "Put on your shoes." The client eventually complies, but the supervisor is concerned about how long the client takes to begin responding after an instruction is delivered. Which measure BEST captures that concern?

2. A client takes bites of food one right after another with almost no pause between them. The supervisor wants a measure of how much time passes between the end of one bite and the beginning of the next. Which measurement procedure should the RBT use?

3. A client's tantrums have not changed in number from week to week, but the RBT notices that some episodes last only seconds while others go on much longer, and the supervisor wants to know whether episodes are becoming shorter. Which measure BEST answers that question?

4. An RBT writes "hand raising: 12" on the data sheet but does not record how long the observation lasted. Why is this a problem?

6. An RBT is told to divide the session into equal intervals and to score an interval as an occurrence if the behaviour happens at any point during that interval. Compared with recording the behaviour continuously, what will these data MOST likely do?

7. A supervisor wants on-task behaviour measured so that an interval is scored as an occurrence only if the client is on task for the entire interval. What should the RBT expect from the resulting data?

8. An RBT is running a group activity and can only glance up briefly at set times. The supervisor asks her to look at the client at the end of each interval and record whether the behaviour is happening at that instant. Which procedure is being described?

9. A supervisor wants an estimate of how much of a session a client spends engaged in an ongoing, continuous behaviour, but the RBT must keep teaching and cannot record every instance. Which sampling method is LEAST likely to systematically overestimate or underestimate that behaviour?

10. Which target behaviour is BEST suited to permanent product recording?

11. An RBT measures bed making by inspecting the bed after the client leaves the room. The parent mentions that she sometimes finishes the bed herself when the client is slow. What is the MAIN measurement concern?

12. An RBT graphs three baseline sessions followed by five intervention sessions on the same line graph. How should the graph be constructed at the point where the intervention began?

13. On a behaviour-analytic line graph, what does a vertical phase-change line tell the reader?

14. Which of the following target definitions is written in observable and measurable terms?

15. An RBT is asked to review four proposed target behaviours. Which one fails the Dead Man's Test and therefore is not written as a behaviour?

16. An RBT records 20 occurrences of hand raising during a 40-minute session. The supervisor asks for the rate per hour. What should the RBT report?

17. During one session a client had 4 tantrum episodes and the recorded durations totalled 12 minutes. What is the mean duration per episode?

18. Aggression is targeted for reduction. Over the last ten sessions the graphed data have moved steadily downward, with the points staying close to that downward path from session to session. How should the RBT describe this data path when reporting to the supervisor?

19. The percentage of correct responses on a skill the client is learning has declined across six consecutive sessions. What is the MOST appropriate action for the RBT?

20. An RBT is very busy during sessions, so instead of recording as the behaviour occurs she writes down her best recollection of the counts at the end of the day. What is the PRIMARY risk of this practice?

21. A fire drill interrupts a session and no data were collected for one of the client's programmes. What should the RBT do?

22. A material required by a written teaching protocol is missing, so an RBT delivers the procedure a different way for the whole session and does not mention it to anyone. What is the MOST significant risk of doing this?