Crossover Multiple Baseline Design
Also known as: CMBD, crossover MBD, multiple baseline crossover design, within-subject multiple baseline design
The crossover multiple baseline design is a single-case experimental design (SCED) that layers crossover sequencing onto a multiple baseline structure. Across two or more tiers — participants, behaviors, or settings — baselines are staggered in time; then treatments are introduced and later reversed or alternated across tiers, so each tier acts as both a treatment and a control unit. The design provides within-subject replication while controlling for time-related confounds.
Key highlights
- Each tier provides both treatment and control data, maximizing causal inference from very small samples.
- Staggered baseline logic rules out maturation, history, and regression-to-mean threats that plague simple pre-post designs.
- Counterbalanced crossover sequencing controls for order effects and allows comparison of two or more conditions within the same study.
- Ethically superior to designs that require prolonged withdrawal of a beneficial treatment, because every tier eventually receives all conditions.
- Replication logic across tiers strengthens external validity even without a large sample.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use this design when you need causal evidence about two or more treatments or conditions for a single individual (or a small group of individuals studied one at a time) and ethical or practical constraints prevent withholding treatment indefinitely. It suits behavioral intervention research, rehabilitation, special education, and clinical psychology where group trials are infeasible. The crossover extension is especially valuable when the researcher wants every tier to experience every condition — maximizing internal comparison power with few participants. Do not use it when the target behavior is irreversible after treatment (learning a skill that cannot be unlearned makes a clean crossover impossible), when strong carryover effects are expected without adequate washout, when baselines cannot be held stable across waiting tiers, or when the research question requires generalizable population-level effect estimates.
Strengths & limitations
- Each tier provides both treatment and control data, maximizing causal inference from very small samples.
- Staggered baseline logic rules out maturation, history, and regression-to-mean threats that plague simple pre-post designs.
- Counterbalanced crossover sequencing controls for order effects and allows comparison of two or more conditions within the same study.
- Ethically superior to designs that require prolonged withdrawal of a beneficial treatment, because every tier eventually receives all conditions.
- Replication logic across tiers strengthens external validity even without a large sample.
- Requires functionally independent tiers; behavioral spillover or social contagion between participants invalidates the waiting-tier control logic.
- Not suitable for irreversible behaviors (skills that, once learned, cannot return to baseline); the crossover assumes some reversibility or comparability of phases.
- Carryover or sequence effects can confound the comparison of conditions if washout periods are insufficient or if the first treatment permanently alters the organism.
- Visual analysis, the standard analytic method, is susceptible to analyst bias; statistical supplements (Tau-U, NAP) improve objectivity but are not universally accepted.
- Small n limits the statistical power of any aggregated effect-size estimates and makes it difficult to generalize findings to a broader population.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How is this different from an alternating treatments design?
Both allow comparison of two or more conditions within a single case, but they differ in structure. An alternating treatments design rapidly switches conditions within the same phase (often within sessions or days), yielding direct comparison of conditions under near-simultaneous circumstances. The crossover multiple baseline design keeps conditions in extended, discrete phases per tier and uses staggered baseline logic to rule out maturation and history threats. The crossover MBD is preferred when you also need the staggered-baseline replication argument and when rapid alternation of conditions is not feasible or appropriate.
How many tiers do I need?
A minimum of three tiers is strongly recommended. With only two tiers, the replication of the effect across waiting baselines is too limited to rule out coincidental timing. Three tiers provide a minimum credible demonstration of experimental control; four or more strengthen it further.
Do I need a washout period between crossover phases?
Whether a washout is needed depends on the nature of the treatment. Pharmacological interventions typically require washout to allow drug clearance. Skill-based behavioral interventions may not require washout if the target behavior is expected to remain reversible, but the researcher should plan and justify the decision a priori based on the theoretical mechanism of the treatment.
Can I use statistical analysis instead of visual analysis?
Visual analysis remains the primary standard in single-case research, as it was built into the design's original logic. However, supplementary statistical indices — such as Tau-U, NAP (non-overlap of all pairs), or IRD (improvement rate difference) — are increasingly recommended to quantify effect size and to support meta-analytic aggregation across studies. They complement rather than replace visual analysis.
Is this the same as an n-of-1 trial?
There is substantial overlap. An n-of-1 trial typically refers to a crossover design applied to a single patient in clinical medicine, often with randomization of condition order and statistical analysis. The crossover multiple baseline design extends across multiple tiers (participants, behaviors, or settings) and uses staggered baselines as a control strategy rather than randomization of period order. The n-of-1 trial is one specific application context; the crossover MBD is a broader SCED framework.
Sources
- 1.Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1(1), 91–97.
- 2.Kazdin, A. E. (2011). Single-Case Research Designs: Methods for Clinical and Applied Settings (2nd ed.). Oxford University Press.ISBN 978-0195341881
You have read it. What now?
Cite this page
ScholarGate. (2026, June 3). Crossover Multiple Baseline Design. ScholarGate. https://scholargate.app/experimental-design/crossover-multiple-baseline-design