Overview
Measurement, Data Display, and Interpretation is Section C of the BCBA outline, with 12 task statements. It runs from writing operational definitions through direct, indirect and product measures, occurrence and temporal measures, continuous and discontinuous recording, efficiency, validity and reliability, graphing and visual analysis, and the measurement of procedural integrity.
The exam treats measurement as a decision rather than a definition. Many items describe a behavior, a goal and a setting, then ask which system will produce valid, feasible data. Others present a graph or a description of phases and ask what the data show. Calculation items appear, such as converting counts to rates or working out interobserver agreement. The 6th edition adds explicit attention to procedural integrity data, which connects to data-based decisions in Section H.
How it's tested
Section C items mix scenario-based selection questions, short calculations and graph interpretation. Expect to read a description of a behavior and setting and pick the best measurement system, or to see described data and decide what they show. Distractors often name a technically correct method that is wrong for this behavior or setting.
-
Choose the system
A behavior, goal and constraint are given, and you pick the measurement procedure that captures the critical dimension feasibly.
-
Predict the error
A discontinuous method is applied to a behavior and you identify whether it will overestimate or underestimate occurrence.
-
Read the graph
Phase data are described and you decide whether level, trend or variability supports a change in conditions.
A teacher wants to increase a student's on-task behavior and can only glance up every two minutes. Options include partial-interval recording, whole-interval recording, momentary time sampling and frequency. Continuous whole-interval observation is impossible with her duties, and partial interval would overestimate on-task behavior. Momentary time sampling fits her constraints and suits a behavior of extended duration, so it is the best choice.
Task statements 12
Official task statements from the Behavior Analyst Certification Board outline, in outline order.
C.1 Create operational definitions of behavior.Writing objective, clear and complete definitions of target behavior.
- What it means
- An operational definition describes a behavior in observable, measurable terms so that independent observers agree on each occurrence. It states what counts, what does not, and often includes examples and nonexamples.
- What to understand
- Good definitions are objective (no inferred states), clear (readable and repeatable) and complete (boundaries for inclusion and exclusion). Function-based definitions group responses by effect; topography-based definitions group them by form. Choose based on the purpose of measurement.
- How it shows up
- Items offer four definitions of a behavior and ask which is best, or ask what is wrong with a given definition.
- Common confusion
- Accepting definitions that include intent or emotion, such as 'gets frustrated'.
C.2 Distinguish among direct, indirect, and product measures of behavior.How measurement relates to the behavior itself.
- What it means
- Direct measurement observes the behavior of interest as it occurs. Indirect measurement relies on reports such as interviews or rating scales. Permanent product measurement records the lasting effects of behavior, such as completed worksheets.
- What to understand
- Direct measurement is preferred where practical. Permanent products are useful when behavior leaves a reliable trace, but the analyst must confirm the target person produced it. Indirect measures add context but carry reporting bias.
- How it shows up
- A scenario describes a data source and asks which type of measurement it represents or which is most appropriate.
- Common confusion
- Treating permanent products as indirect measures.
C.3 Measure occurrence.Count, frequency and rate, and related occurrence measures.
- What it means
- Occurrence measures include count, rate or frequency (count per unit of time), and percentage of opportunities. Rate allows comparison across observation periods of different lengths.
- What to understand
- Use rate when observation lengths vary and the behavior is free operant. Use percentage or trials to criterion for discrete trials with set opportunities. Count alone is misleading when session lengths differ.
- How it shows up
- Expect calculation items such as converting counts from sessions of different durations into rates.
- Common confusion
- Comparing raw counts across sessions of unequal length.
C.4 Measure temporal dimensions of behavior (e.g., duration, latency, interresponse time).How long behavior lasts and when it occurs.
- What it means
- Duration measures how long a response lasts. Latency measures the time from a stimulus to the start of a response. Interresponse time (IRT) measures the time between the end of one response and the start of the next.
- What to understand
- Choose duration when the concern is how long behavior lasts, such as tantrums or on-task time. Choose latency for compliance or response initiation. Choose IRT when the spacing of responses matters, as with rapid eating.
- How it shows up
- A vignette describes a problem, such as slow starts after instructions, and asks which temporal measure fits.
- Common confusion
- Confusing latency (stimulus to response) with IRT (response to response).
C.5 Distinguish between continuous and discontinuous measurement procedures.Recording every instance versus sampling behavior.
- What it means
- Continuous measurement detects every instance of the behavior during the observation period, as with count, duration or latency recording. Discontinuous measurement samples behavior at intervals and estimates its occurrence.
- What to understand
- Continuous measurement gives the most accurate picture but can be impractical. Discontinuous methods are efficient but introduce known errors, so they must be selected with care for each behavior and goal.
- How it shows up
- Items ask which method captures all instances of a behavior or why a sampling method misrepresents behavior.
- Common confusion
- Calling momentary time sampling continuous because it runs through the whole session.
C.6 Design and apply discontinuous measurement procedures (e.g., interval recording, time sampling).Whole interval, partial interval and momentary time sampling.
- What it means
- Whole-interval recording scores an interval only if behavior occurs throughout it. Partial-interval recording scores it if behavior occurs at any point. Momentary time sampling scores it if behavior occurs at the moment the interval ends.
- What to understand
- Whole interval tends to underestimate behavior and suits behavior to increase. Partial interval tends to overestimate and suits behavior to decrease. Momentary time sampling can over- or underestimate and suits longer-duration behavior; planned activity check applies it to groups.
- How it shows up
- Items ask which method to select for a behavior to increase or decrease, or which error a method will produce.
- Common confusion
- Using partial interval for a behavior the team wants to increase, which inflates progress.
C.7 Measure efficiency (e.g., trials to criterion, cost-benefit analysis, training duration).How quickly and economically outcomes are achieved.
- What it means
- Efficiency measures compare how much instruction, time or resources are needed to reach an outcome. Examples include trials to criterion, sessions or minutes to mastery, and cost-benefit comparisons of alternative procedures.
- What to understand
- Two effective procedures may differ greatly in efficiency. Efficiency data help choose between interventions and justify resource use to funders and families.
- How it shows up
- A scenario compares two teaching methods with similar outcomes and asks which measure would show the better choice.
- Common confusion
- Assuming the procedure with the higher final accuracy is automatically more efficient.
C.8 Evaluate the validity and reliability of measurement procedures.Whether measurement captures the right thing consistently.
- What it means
- Validity means measuring the socially significant behavior directly, with a relevant dimension, at times and places that represent it. Reliability means the same events yield the same values when measured repeatedly. Accuracy means the data match the true values.
- What to understand
- Interobserver agreement (IOA) shows that observers agree, which supports believability but does not prove accuracy. Threats include poorly designed systems, observer drift, reactivity and expectations. Calibration against a true value is the test of accuracy.
- How it shows up
- Items describe declining agreement over time and ask what threat is likely, or ask which IOA method suits a data type.
- Common confusion
- Equating high IOA with accuracy.
C.9 Select a measurement procedure to obtain representative data that accounts for the critical dimension of the behavior and environmental constraints.Matching measurement to behavior and setting.
- What it means
- This task asks you to choose a measurement system based on the dimension of behavior that matters (such as rate, duration or latency) and on practical limits such as staff availability, setting and the observer's other duties.
- What to understand
- Start with the clinical question, then pick the dimension that reflects it, then choose the most accurate method the setting can support. Where a discontinuous method is required, pick one whose error does not bias the conclusion.
- How it shows up
- A busy teacher with few free minutes needs data on a behavior, and you select the most valid feasible method.
- Common confusion
- Choosing the most precise system even when staff cannot implement it reliably.
C.10 Graph data to communicate relevant quantitative relations (e.g., equal-interval graphs, bar graphs, cumulative records).Choosing and constructing appropriate graphs.
- What it means
- Line graphs on equal-interval axes are the standard for showing change over time. Bar graphs summarise discrete categories or conditions. Cumulative records show total responses over time, where slope indicates rate. Standard celeration charts plot rate on a logarithmic scale.
- What to understand
- Graphs should include labeled axes, condition change lines, condition labels and data paths broken between conditions. The choice of graph depends on what relation you need to communicate.
- How it shows up
- Items ask which graph best shows a relation, or what a steep slope means on a cumulative record.
- Common confusion
- Reading a flat line on a cumulative record as zero responding overall rather than no new responses.
C.11 Interpret graphed data.Visual analysis of level, trend and variability.
- What it means
- Visual analysis examines level, trend and variability within and across phases, along with immediacy of effect and overlap between conditions. Decisions rest on stable patterns rather than single data points.
- What to understand
- Look for a stable baseline before changing conditions. A change in level or trend that follows the condition change and is replicated supports a functional relation. High variability calls for more data or a search for uncontrolled variables.
- How it shows up
- A graph or description of phase data is given and you judge whether the intervention produced a change.
- Common confusion
- Concluding effectiveness when baseline already showed a trend in the therapeutic direction.
C.12 Select a measurement procedure to obtain representative procedural integrity data that accounts for relevant dimensions (e.g., accuracy, dosage) and environmental constraints.Measuring whether an intervention is implemented as planned.
- What it means
- Procedural integrity, also called treatment fidelity, measures how accurately and how completely an intervention is implemented. Dimensions include accuracy of each step and dosage, meaning how often and how long the intervention is delivered.
- What to understand
- Without integrity data, poor outcomes cannot be attributed to the plan or to its implementation. Task analysis checklists and direct observation are common methods. Section H.6 uses these data to make decisions.
- How it shows up
- A plan shows no effect, and you choose which data to collect before changing the intervention.
- Common confusion
- Changing the intervention before checking whether it was implemented correctly.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
Operational definition
Observable, measurable description with boundaries.
- Remember
- Objective, clear and complete.
- Tested as
- Best-definition items appear regularly.
- Trap
- Including intent or feelings.
-
Rate
Count divided by observation time.
- Remember
- Use when session lengths differ.
- Tested as
- Calculation items.
- Trap
- Comparing raw counts across uneven sessions.
-
Latency vs IRT
Stimulus-to-response time versus response-to-response time.
- Remember
- Latency suits compliance; IRT suits pacing.
- Tested as
- Selection and definition items.
- Trap
- Swapping the two.
-
Whole-interval recording
Scores only if behavior lasts the entire interval.
- Remember
- Underestimates; use for behavior to increase.
- Tested as
- Error-direction items.
- Trap
- Using it for behavior to decrease.
-
Partial-interval recording
Scores if behavior occurs at any point.
- Remember
- Overestimates; use for behavior to decrease.
- Tested as
- Error-direction items.
- Trap
- Using it for behavior to increase.
-
Momentary time sampling
Scores behavior at the end of each interval only.
- Remember
- Suits longer-duration behavior; error can go either way.
- Tested as
- Feasibility items in classrooms.
- Trap
- Using it for brief, low-rate behavior.
-
Interobserver agreement
Agreement between independent observers.
- Remember
- Supports believability, not proof of accuracy.
- Tested as
- IOA method and threat items.
- Trap
- Treating IOA as accuracy.
-
Observer drift
Unplanned change in how an observer applies a definition.
- Remember
- Prevent with retraining and periodic IOA.
- Tested as
- Threat identification items.
- Trap
- Blaming the client's behavior for the change.
-
Reactivity
Behavior changes because it is being observed.
- Remember
- Reduce with unobtrusive measurement.
- Tested as
- Validity threat items.
- Trap
- Confusing it with observer drift.
-
Level, trend, variability
Core dimensions of visual analysis.
- Remember
- Need stable baseline before changing phases.
- Tested as
- Graph interpretation items.
- Trap
- Judging from the last data point.
-
Cumulative record
Total responses plotted over time.
- Remember
- Slope shows rate; flat means no new responses.
- Tested as
- Graph reading items.
- Trap
- Reading a flat segment as zero total.
-
Procedural integrity
Accuracy and dosage of implementation.
- Remember
- Check before judging effectiveness.
- Tested as
- New emphasis in the 6th edition.
- Trap
- Changing the plan without integrity data.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
Discontinuous measurement methods
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Whole interval | Behavior occurs throughout the interval. | Underestimates occurrence. | Selecting it for behavior to reduce. |
| Partial interval | Behavior occurs at any time in the interval. | Overestimates occurrence. | Selecting it for behavior to increase. |
| Momentary time sampling | Behavior occurs at the end of the interval. | Bias depends on behavior pattern. | Using it for brief, rare behavior. |
Temporal measures
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Duration | How long a response lasts. | Measures the response itself. | Using it for discrete short responses. |
| Latency | Time from stimulus to response onset. | Starts with an antecedent. | Confusing with IRT. |
| Interresponse time | Time between consecutive responses. | Starts with the previous response. | Calling it latency. |
Measurement quality
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Validity | Measures the right behavior and dimension. | About what is measured. | Assuming reliable data are valid. |
| Reliability | Consistent results on repeated measurement. | About consistency. | Assuming consistency equals truth. |
| Accuracy | Data match true values. | Requires comparison with a true value. | Using IOA as evidence of accuracy. |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
Most precise method wins
- What candidates pick
- Continuous duration recording
- Why it looks right
- Continuous methods are the most accurate.
- Why it's wrong
- If the setting cannot support it, the data will be incomplete or unreliable; feasibility is part of the task.
Clue to watch for The stem mentions staff limits or other duties.
-
Partial interval for a skill
- What candidates pick
- Partial interval
- Why it looks right
- It is easy to apply.
- Why it's wrong
- Partial interval inflates occurrence, so progress on a behavior to increase looks better than it is.
Clue to watch for The goal is to increase the behavior.
-
IOA proves accuracy
- What candidates pick
- The data are accurate
- Why it looks right
- Two observers agreed closely.
- Why it's wrong
- Both observers can be wrong in the same way; accuracy requires comparison with true values.
Clue to watch for Look for a calibration or true-value standard.
-
Latency vs IRT
- What candidates pick
- Interresponse time
- Why it looks right
- Both measure time before a response.
- Why it's wrong
- If timing starts at an instruction or other stimulus, it is latency.
Clue to watch for The clock starts when the instruction is given.
-
Trend in baseline ignored
- What candidates pick
- The intervention worked
- Why it looks right
- Treatment-phase data look better.
- Why it's wrong
- A baseline already improving means the change cannot be attributed to the intervention.
Clue to watch for Baseline data slope in the therapeutic direction.
-
Skip integrity checks
- What candidates pick
- Modify the intervention
- Why it looks right
- Outcome data look poor.
- Why it's wrong
- Poor results may reflect poor implementation; integrity data are needed first.
Clue to watch for The stem mentions new staff or inconsistent implementation.
Answer framework
Apply these steps when selecting a measurement system or interpreting described data.
- State the clinical question first, then identify the dimension of behavior that answers it, such as rate, duration, latency or interresponse time.
- Decide whether continuous measurement is feasible in this setting, given the number of staff, the time available and the observer's other duties.
- If sampling is needed, pick the discontinuous method whose known bias will not exaggerate progress toward the goal the team has set.
- Plan interobserver agreement and procedural integrity checks so the data can be believed and any outcome can be attributed to the plan.
- When reading data, examine level, trend and variability within and across phases before drawing any conclusion about the intervention's effect.
Practice question
Want more like this? The BCBA practice test has questions for every domain.
Checklist
0 of 12
Tick each one you can do without notes. Your ticks stay in this browser.
Study strategy
Plan 25 to 35 hours for Section C. Start with operational definitions and the interval methods, because they feed many scenario items. Build a one-page table that shows each measurement method, what it captures, its bias and when to use it, then quiz yourself by reading short behavior descriptions and choosing a method within 30 seconds.
Next, practise calculations: rate, percentage of opportunities, and the common IOA formulas such as total count, exact count-per-interval and interval-by-interval. Work several problems by hand so the arithmetic becomes quick and automatic. Then spend time on visual analysis. Use published single-case graphs from journal articles and describe level, trend and variability in each phase before reading the authors' conclusions.
Finish with procedural integrity. Write a short integrity checklist for an intervention you know, including accuracy and dosage, and decide how often to collect it. Revisit Section C alongside Section D, since design questions depend on reading data correctly.
FAQ
How much of the BCBA exam covers Measurement, Data Display, and Interpretation?
Measurement, Data Display, and Interpretation is Section C and has 21 of the 175 scored questions, about 12% of the exam.
Are there calculation questions in this section?
Yes. Expect items on rate, percentage, and interobserver agreement. Practise the formulas by hand until each takes well under a minute.
Which interval method should I use for a behavior to increase?
Whole-interval recording is usually preferred for behavior to increase, because its underestimation will not exaggerate progress. Momentary time sampling suits longer-duration behavior when observers have other duties.
Does interobserver agreement show the data are accurate?
No. IOA shows observers agreed. Accuracy requires comparing data with true values, for example from a calibrated recording.
What is new about procedural integrity in the 6th edition?
Task C.12 asks you to select a procedure for measuring integrity that captures accuracy and dosage within real constraints. It pairs with H.6, which asks you to act on those data.
How should I practise graph interpretation?
Read published single-case graphs and describe level, trend and variability before reading the authors' conclusions. Then compare your judgement with theirs.
Test your Domain 3 knowledge
Practise choosing measurement systems, calculating interobserver agreement and reading phase data with timed items modelled on Section C.