BCBA Experimental Design Questions With Worked Reasoning

If you are mapping the BCBA exam for the first time, start with the BACB sixth-edition Test Content Outline. It breaks down the 6th edition Test Content Outline and shows where Area D fits. When you are ready for more practice, move to the free questions at /free-practice/bcba.

What experimental control means

Experimental control is the heart of behavior-analytic research. It means you can show that a change in the dependent variable was caused by the independent variable, not by something else that happened at the same time.

In single-subject research, control is built on three ideas:

• Prediction: During baseline, behavior is stable enough that you can predict what it will look like if conditions stay the same.
• Verification: When the intervention is withdrawn or when it is introduced at a later time, the earlier baseline level returns or is reproduced.
• Replication: Reintroducing the intervention produces the effect again, strengthening the claim that the intervention — not coincidence — caused the change.

A functional relation is stronger when the effect is large, immediate, replicated across phases or baselines, and unlikely to be explained by outside events.

Single-subject designs the exam expects

Reversal / withdrawal design

In an A-B-A-B design, the researcher measures baseline (A), introduces the intervention (B), withdraws it to return to baseline conditions (A), and then reintroduces the intervention (B). The logic is simple: if behavior changes each time the independent variable is added or removed, the variable is probably responsible.

Strength: It provides a strong demonstration of a functional relation because prediction, verification, and replication are all visible.

Limit: It cannot be used when the behavior would not return to baseline — for example, after a skill has been learned — or when withdrawal would be unsafe or unethical.

Multiple baseline design

A multiple-baseline design introduces the same intervention at staggered times across behaviors, settings, or subjects. Each baseline stays untreated until its scheduled turn. If behavior changes only when the intervention reaches that specific baseline, the design argues against history or maturation as the cause — those outside events would have affected all baselines at once.

Strength: No withdrawal is needed, so it works for acquired skills and ethical situations where reversal is impossible.

Limit: The baselines must be independent. If changing one behavior also changes another, the design cannot support a clean causal claim.

Alternating treatments design

An alternating-treatments design compares two or more conditions by rapidly alternating them across sessions or within the same session. For example, a learner might receive Treatment X in the morning and Treatment Y in the afternoon, with the order randomized. The design shows which condition produces better performance, not necessarily that either condition caused behavior to change from baseline.

Strength: It is efficient and does not require baseline, withdrawal, or staggered implementation.

Limit: Carryover, sequence, or contrast effects can distort results, so condition discrimination, counterbalancing, measurement quality, and repeated differentiation matter. A well-designed comparison can demonstrate experimental control; there is no universal ranking of these designs.

Changing-criterion design

A changing-criterion design sets a series of stepwise performance goals. The intervention is applied throughout, and the criterion is raised only after behavior stabilizes at the current level. If the participant meets each new criterion as it is set, the researcher argues that the intervention is controlling behavior.

Strength: It is useful for shaping, academic or athletic goals, and any situation where behavior must move gradually in one direction.

Limit: It is vulnerable to history effects if many criterion changes occur on the same schedule, and it requires the behavior to track each step precisely.

Competing explanations

Even when behavior changes after an intervention, other explanations may be possible. The exam tests whether you can spot the most plausible competing explanation and whether the design used rules it out.

Maturation

Maturation means the participant changed naturally over time — development, recovery, habituation, or fatigue. A design rules out maturation when behavior changes immediately with the introduction or withdrawal of the independent variable, and when the effect replicates across phases or baselines. Natural maturation would not reverse when the intervention is withdrawn.

History

History means an outside event happened at the same time as the intervention: a new teacher, a medication change, a family move, or a school policy change. Reversal designs rule out simple history because the outside event would not disappear when the intervention is withdrawn. Multiple-baseline designs rule out history when effects occur only at staggered introduction times, not all at once.

Measurement artifacts

Measurement artifacts include instrumentation changes (a new data sheet, observer, or definition), observer drift or reactivity, and regression to the mean. Good studies use clear operational definitions, observer training, and interobserver agreement checks to keep measurement consistent. A sudden change that coincides with a new observer or a new definition should raise suspicion.

Worked questions

Each scenario is an original four-option item. The correct answer is followed by a rationale for every option.

Question 1

A behavior analyst measures non-dangerous calling out during a supervised classroom activity. After five stable baseline sessions at 8–10 episodes, she introduces a peer-buddy intervention. Calling out falls to 1–2 episodes per session. She removes the peer buddy for three sessions and calling out returns to 8–9 episodes. When she reintroduces the peer buddy, calling out falls again.

Which design best describes this study?

1. Reversal / withdrawal design
2. Multiple baseline across subjects
3. Alternating treatments design
4. Changing-criterion design

Question 2

A BCBA wants to compare a new token economy with the existing level system for reducing disruptive behavior in a classroom. Both procedures are appropriate for this learner, and a safety review permits brief comparisons. The procedures have distinct cues and minimal expected carryover; returning to untreated baseline is not planned. The BCBA needs a comparison within the same week.

Which design is most appropriate?

1. Reversal / withdrawal design
2. Multiple baseline across settings
3. Alternating treatments design
4. Changing-criterion design

Question 3

During baseline, a learner's self-injury is stable at 12–15 episodes per day. The BCBA starts a noncontingent reinforcement procedure. Self-injury drops to 2–3 episodes per day. At the same time, the learner's medication is increased and a new communication device is introduced.

What is the most accurate conclusion?

1. Experimental control is demonstrated because the behavior changed after the intervention.
2. Control is questionable because history and multiple-treatment confounds are present.
3. Control is not shown because no reversal was conducted.
4. Control is not shown because of maturation.

Question 4

A supervisor trains three RBTs to implement a prompting protocol. Baseline data are collected for each RBT. Training is introduced first for RBT 1, then later for RBT 2, and finally for RBT 3. Each RBT's correct implementation increases only after their own training begins.

Which design best describes this study?

1. Multiple baseline across behaviors
2. Multiple baseline across subjects
3. Alternating treatments design
4. Changing-criterion design

Question 5

A school-wide positive-behavior intervention is associated with a decrease in office referrals over a semester. During the same semester, the school also hires a new principal and changes the discipline policy.

Which competing explanation is most important?

1. Maturation
2. History
3. Regression toward the mean
4. Instrumentation

Question 6

A BCBA is working with an adult client to increase daily steps. The activity plan is medically cleared, and the client agrees to gradual increases within that plan. The BCBA sets weekly step goals that increase by 500 steps each week and provides reinforcement when the client meets the current week's criterion.

Which design is most appropriate?

1. Reversal / withdrawal design
2. Multiple baseline across settings
3. Alternating treatments design
4. Changing-criterion design

Question 7

A BCBA alternates 10-minute sessions of two attention-based interventions in random order each day. No baseline is collected. The analyst compares the frequency of aggression under each schedule.

Which statement is most accurate?

1. This is a reversal design.
2. This is a multiple-baseline design.
3. This is an alternating-treatments design that compares two procedures but has no baseline.
4. This is a changing-criterion design.

What to practice next

If these scenarios exposed a gap, go back to the definitions of each design and ask yourself: Does the study manipulate the independent variable, can the effect be verified, and could an outside event explain the change? That three-part question is the fastest way to classify BCBA experimental-design items.

For more BCBA reasoning practice, visit /free-practice/bcba. For the full exam map, see the BACB sixth-edition Test Content Outline.