Crossover Single-Subject Experimental Design
Also known as: crossover SSED, alternating-treatments crossover design, single-case crossover design, N-of-1 crossover design
The crossover single-subject experimental design (crossover SSED) applies two or more treatment conditions sequentially to the same individual, with a washout or return-to-baseline period between conditions. Because each participant serves as their own control, between-subject variability is eliminated, enabling precise causal inference about treatment effects even with very small samples — often a single participant. This design is widely used in applied behavior analysis, special education, rehabilitation, and clinical psychology.
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 a crossover SSED when you need to compare two or more treatments within the same individual and sufficient participants for a group RCT are unavailable or ethically problematic. It is ideal in rare-condition clinical research, special education, rehabilitation, and applied behavior analysis where individualized causal inference is more valuable than population-level estimation. It requires that the outcome be repeatedly measurable and that treatment effects are not expected to be permanent (i.e., some reversibility or washout is plausible). Do NOT use this design if carryover effects cannot be plausibly eliminated between phases — for example when the intervention teaches a durable skill that cannot be unlearned — or when ethical concerns preclude withdrawing an effective treatment.
Strengths & limitations
- Eliminates between-participant variability by using each individual as their own control, maximizing internal causal validity.
- Enables rigorous experimental inference with very small samples (even N=1), making it feasible for rare conditions and low-incidence populations.
- Intensive repeated measurement yields detailed information about within-participant response patterns and variability across time.
- Well-suited to individualized or person-centered research where population averages obscure clinically meaningful individual differences.
- Results from multiple single-subject crossover studies can be aggregated through single-case meta-analysis to build broader evidence.
- External generalizability is limited; findings describe the specific individual(s) studied and cannot be statistically extrapolated to a population without replication across participants.
- Carryover effects — where the first treatment's impact persists into the second phase — can confound the comparison if the washout period is insufficient.
- Order effects (the sequence in which conditions are presented) may interact with treatment, and a single-participant design cannot randomize order across participants.
- Demands consistent, intensive measurement over extended periods, placing burden on both researcher and participant and raising the risk of attrition or protocol drift.
Frequently asked
How is a crossover SSED different from an ABA reversal design?
Both use the same participant across phases, but their logic differs. An ABA reversal design withdraws a single treatment to demonstrate experimental control by returning to baseline (A-B-A or A-B-A-B). A crossover SSED sequentially applies two distinct treatment conditions (B and C) with a washout between them to directly compare which treatment produces a larger effect. The crossover design answers 'which treatment works better?'; the reversal design answers 'does this treatment cause the change?'.
What is an adequate washout period?
The washout length must be theoretically justified based on how long the first treatment's effects are expected to persist. For behavioral interventions targeting specific discrete behaviors, returning to stable baseline data (typically 3–5 stable or counter-therapeutic data points) is the standard. For pharmacological N-of-1 crossover trials, washout is typically defined as five or more half-lives of the drug. There is no universal number of sessions; the criterion is demonstrated return to pre-treatment performance levels.
Can I use this design with more than two treatment conditions?
Yes. The design extends naturally to three or more conditions (e.g., A-B-washout-C-washout-D), though each additional phase lengthens the study and increases the risk of participant attrition, sequence-by-treatment interactions, and cumulative carryover. With three or more conditions, counterbalancing order across a small number of participants (if feasible) helps control sequence effects.
How do I report effect sizes for a crossover SSED?
The most widely recommended non-overlap effect sizes for single-case data are Tau-U (which adjusts for baseline trend) and Non-overlap of All Pairs (NAP). Both can be computed phase-by-phase and aggregated across replications. Parker et al. (2011) provide accessible formulas and benchmarks for Tau-U. Reporting at least one validated effect-size index alongside visual analysis is increasingly required by journals and What Works Clearinghouse standards.
Is a crossover SSED sufficient for evidence-based practice claims?
Single-subject research, including crossover SSEDs, is recognized as a legitimate experimental design for evidence-based practice in special education, behavior analysis, and clinical psychology when replication criteria are met — typically a minimum of three demonstrations of effect across different participants, settings, or time points. A single crossover study on one participant alone does not establish an evidence base, but systematic replication across participants does.
Sources
- Kazdin, A. E. (2011). Single-Case Research Designs: Methods for Clinical and Applied Settings (2nd ed.). Oxford University Press. ISBN: 978-0195341881
- Barlow, D. H., Nock, M. K., & Hersen, M. (2009). Single Case Experimental Designs: Strategies for Studying Behavior Change (3rd ed.). Pearson. ISBN: 978-0205474554
How to cite this page
ScholarGate. (2026, June 3). Crossover Single-Subject Experimental Design. ScholarGate. https://scholargate.app/en/experimental-design/crossover-single-subject-experimental-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
- N-of-1 TrialClinical Research↔ compare
- Randomized Controlled TrialExperimental design↔ compare
- Single-Subject Experimental DesignExperimental design↔ compare