Crossover ABAB Design — ABAB Reversal Design
Crossover ABAB Reversal Design · Also known as: ABAB reversal design, reversal design, withdrawal design, ABAB single-subject design
The crossover ABAB design is a single-subject experimental design that alternates between baseline (A) and intervention (B) conditions twice within the same participant. By withdrawing and reintroducing the treatment, the researcher can demonstrate experimental control: if behavior improves with B and reverts with A, the causal link between the intervention and the outcome is established without a separate control group.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use the ABAB design when you need to establish a causal relationship between a specific intervention and a behavioral or psychological outcome in an individual or very small sample, and when it is ethically and practically feasible to withdraw the treatment temporarily. It is well suited to applied behavior analysis, special education, rehabilitation, and clinical psychology. Do not use it when withholding treatment during A2 would expose the participant to harm, when the target behavior is irreversible once learned (e.g., acquisition of a new skill that will not be forgotten), or when carryover effects make a clean reversal implausible.
Strengths & limitations
- Establishes experimental control and causal inference within a single participant, eliminating between-subject variability.
- The internal replication across two AB sequences greatly strengthens confidence compared to a simple AB design.
- Requires no separate control group, making it feasible for rare populations or individual clinical cases.
- Continuous measurement provides rich time-series data that reveal the trajectory and immediacy of change.
- Flexible length per phase — each phase continues until stability criteria are met rather than being fixed in advance.
- Requires that the target behavior is reversible when the intervention is withdrawn; irreversible skills cannot be studied with this design.
- Ethical concerns arise whenever withholding a beneficial treatment (phase A2) exposes the participant to harm or distress.
- External generalizability is limited; findings apply first and foremost to the individual studied.
- Sequential confounds such as order effects, fatigue, or historical events can threaten internal validity if phases are long.
- Carryover or contrast effects between adjacent phases may obscure the true magnitude of the treatment effect.
Frequently asked
What is the difference between a crossover ABAB design and a simple AB design?
An AB design applies baseline then intervention just once and cannot rule out the possibility that the change was caused by something other than the treatment (a history effect, maturation, regression to the mean). The ABAB design adds a withdrawal phase (A2) and a second treatment phase (B2). If behavior reverses in A2 and improves again in B2, the repeated co-variation provides strong causal evidence — a built-in replication that the simple AB design lacks.
How many data points are needed in each phase?
There is no fixed minimum prescribed by law, but applied behavior analysis convention requires at least three data points per phase before transitioning, and more if the data show instability or trend. In practice, phases often last five to ten measurement occasions to establish a reliable estimate of the behavior's level and trend under each condition.
Is it ethical to withdraw a working treatment in the A2 phase?
This is the central ethical tension of the design. The withdrawal must be pre-approved by an ethics board, explained to participants and caregivers, and accompanied by a stopping rule — a criterion for early reintroduction if the participant's welfare is compromised. For behaviors that pose safety risks, withdrawal designs are generally contraindicated, and a multiple-baseline design is preferred instead.
How is ABAB different from a crossover randomized trial?
A crossover randomized controlled trial (RCT) also exposes participants to multiple conditions, but it involves a group of participants, randomizes the order of conditions, and uses statistical models (mixed effects, carryover adjustment) for analysis. ABAB is a single-subject design with sequential rather than randomized phase order, analyzed primarily through visual inspection and non-overlap statistics. ABAB is more common in behavior analysis; crossover RCTs are more common in clinical medicine.
Can I extend the design beyond ABAB?
Yes. Researchers sometimes use ABABAB or longer sequences to strengthen the replication evidence or to test a variant of the intervention. Additionally, designs can be extended to ABACABAC formats when comparing two treatments (B and C) against baseline. The principle remains the same: systematic alternation with measurement at every point.
Sources
- Barlow, D. H., Nock, M. K., & Hersen, M. (2009). Single Case Experimental Designs: Strategies for Studying Behavior Change (3rd ed.). Pearson. ISBN: 978-0205474929
- Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied Behavior Analysis (3rd ed.). Pearson. ISBN: 978-0134752556
How to cite this page
ScholarGate. (2026, June 3). Crossover ABAB Reversal Design. ScholarGate. https://scholargate.app/en/experimental-design/crossover-abab-design
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Alternating Treatments DesignDisability Studies↔ compare
- Multiple Baseline DesignExperimental design↔ compare
- Randomized Controlled TrialExperimental design↔ compare