What Is a Reversal Design?
The reversal design is one of the most commonly tested research methodologies on the BCBA exam. It is a single-subject experimental design that demonstrates functional relationships between an independent variable (intervention) and a dependent variable (behavior) by alternating between conditions. The most common form is the ABAB design, where A represents baseline and B represents the treatment condition.
The power of the reversal design lies in its ability to demonstrate experimental control clearly. If behavior changes when treatment is introduced, returns toward baseline levels when treatment is removed, and changes again when treatment is reintroduced, there is strong evidence that the intervention caused the behavior change.
Phases of the ABAB Design
Phase 1: First Baseline (A1)
During the initial baseline, the researcher measures the target behavior without any intervention in place. This phase serves a critical purpose: it establishes the pattern of responding that would be expected to continue without treatment. The baseline should be continued until the data show a stable pattern or a clear trend. An unstable or highly variable baseline makes it difficult to detect changes when the intervention is introduced.
Data collection during baseline should use the same measurement procedures that will be used during treatment. Consistency in measurement across all phases is essential for valid comparisons.
Phase 2: First Treatment (B1)
The intervention is introduced while data collection continues. The researcher looks for changes in level, trend, or variability compared to baseline. A clear and immediate change in the data path provides the strongest evidence that the intervention is responsible for the behavior change.
The treatment phase should continue until the behavior reaches a stable level at the desired criterion or shows a clear change from baseline patterns.
Phase 3: Return to Baseline (A2)
The intervention is withdrawn, and conditions return to those in the original baseline. This is where the term "reversal" comes from. If the behavior returns toward baseline levels, this provides verification that the treatment was responsible for the change observed in B1. This phase is essential for ruling out alternative explanations such as maturation or coincidental environmental changes.
Some clinicians and researchers struggle with this phase because it requires temporarily removing an effective treatment. This ethical tension is an important consideration when selecting a research design.
Phase 4: Second Treatment (B2)
The intervention is reintroduced. If the behavior improves again, this constitutes replication of the treatment effect. With prediction (A1), verification (A2), and replication (B2), the reversal design provides a compelling demonstration of experimental control.
Withdrawal vs. Reversal: An Important Distinction
Although the terms are often used interchangeably, there is a technical distinction between withdrawal and reversal designs. In a withdrawal design, the treatment is simply removed during the return-to-baseline phase. In a reversal design, the treatment contingencies are reversed. For example, if the treatment involved reinforcing appropriate behavior, a true reversal would involve reinforcing the inappropriate behavior during the reversal phase.
In practice, true reversals are rarely used because they involve actively reinforcing undesirable behavior. Most designs called "reversal designs" are technically withdrawal designs. However, on the BCBA exam, the terms are typically used interchangeably unless the question specifically asks about the distinction.
When to Use a Reversal Design
The reversal design is appropriate when several conditions are met. First, the behavior must be expected to reverse when the intervention is withdrawn. Behaviors maintained by environmental contingencies (such as attention or escape) are good candidates. Behaviors that, once acquired, are unlikely to return to baseline levels (such as reading skills) are poor candidates.
Second, it must be ethically acceptable to withdraw the intervention temporarily. If the behavior is dangerous (such as severe self-injury) or if withdrawal would cause significant harm, a different design should be selected.
Third, the behavior should be relatively sensitive to environmental changes so that effects can be observed within a reasonable timeframe.
Ethical Considerations
The primary ethical concern with reversal designs involves the temporary withdrawal of an effective treatment. The BACB Ethics Code requires behavior analysts to prioritize client welfare. Several strategies can mitigate ethical concerns:
- Keep reversal phases brief. Only maintain the return-to-baseline phase long enough to observe a trend toward previous levels.
- Use safety criteria. Establish predetermined criteria for ending the reversal phase early if the behavior reaches dangerous levels.
- Consider alternative designs. If withdrawal poses unacceptable risk, a multiple baseline design may be more appropriate.
- Obtain informed consent. Ensure clients and caregivers understand and agree to the temporary withdrawal of treatment.
Variations of the Reversal Design
Beyond the basic ABAB format, several variations exist. The BAB design begins with a treatment phase, which may be appropriate when baseline data are not feasible to collect before treatment begins. The ABAB with multiple treatments (ABACBC) can compare two different treatments, though this requires careful interpretation. Extended designs (ABABAB) with additional phases can strengthen the demonstration of control.
Common Exam Questions
On the BCBA exam, expect questions asking you to identify when a reversal design is appropriate, recognize threats to internal validity in a reversal design scenario, and distinguish between reversal and withdrawal designs. You may also see questions about what to do when a behavior does not reverse during the return-to-baseline phase, which could indicate the behavior has been learned and is now maintained by naturally occurring contingencies.
Understanding the reversal design thoroughly, including its logic, phases, limitations, and ethical implications, is essential for success on both the exam and in clinical practice.