Crossover Laboratory Experiment — Within-Subjects Controlled Design
Crossover Within-Subjects Laboratory Experiment · Also known as: within-subjects crossover lab study, repeated-measures crossover experiment, crossover controlled lab experiment, within-person laboratory crossover trial
A crossover laboratory experiment is a within-subjects experimental design conducted in a controlled lab environment in which each participant receives two or more treatments sequentially, serving as their own control. By eliminating between-person variability from the error term, it yields high statistical power with relatively small samples. Treatment order is randomized or counterbalanced across participants to guard against order and carryover effects.
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 laboratory experiment when the research question concerns the effect of two or more treatments on individuals and the following conditions hold: (1) the outcome is expected to return to a stable baseline after each treatment (washout is achievable), (2) the sample is small or recruitment is costly and within-person power gains are essential, and (3) tight experimental control over stimulus delivery and measurement is feasible or required. It is especially valuable in pharmacology, cognitive psychology, exercise physiology, and human factors research. Do not use it when carryover effects cannot be adequately washed out, when the treatment permanently alters participants (learning or conditioning that does not decay), when the study duration required for multiple sessions creates unacceptable attrition, or when only one treatment can ethically be administered per person.
Strengths & limitations
- Each participant acts as their own control, eliminating between-subject variability from the error term and dramatically increasing statistical power.
- Requires substantially fewer participants than a parallel-group design to achieve the same power — a major advantage when recruitment is difficult or expensive.
- The controlled laboratory environment ensures high internal validity: stimulus delivery, measurement, and conditions are standardized across all treatment periods.
- Allows direct within-person estimation of treatment differences, which is often the theoretically meaningful quantity in cognitive, physiological, and behavioral studies.
- Randomized treatment-order assignment guards against systematic period and order effects at the group level.
- Carryover effects — residual influence of an earlier treatment on a later period — can bias treatment comparisons if the washout interval is insufficient.
- Period effects (e.g., learning, fatigue, or seasonal variation) may confound treatment estimates even with randomized order.
- Requires participants to commit to multiple laboratory sessions, increasing dropout risk and logistical complexity.
- Not appropriate when treatments permanently alter the participant (e.g., surgery, irreversible learning) or when stable baseline cannot be recovered.
- Analysis is more complex than a simple two-group comparison and requires careful modeling of the repeated-measures and sequence structure.
Frequently asked
How long should the washout period be?
The washout interval should be at least five half-lives of the treatment effect — whether pharmacological, physiological, or psychological. For drug studies this is derived from pharmacokinetic data. For behavioral or cognitive treatments, pilot data or literature on the duration of practice or priming effects should guide the choice. Always measure a baseline at the start of each period to verify that the previous treatment's influence has dissipated.
Can I use a crossover design with more than two treatments?
Yes. With three or more treatments, a Latin square or Williams design assigns sequences such that each treatment appears equally often in each period and each treatment precedes every other treatment at least once. This balances first-order carryover across conditions. The required sample size grows with the number of treatment sequences, but the within-subject power advantage is retained.
What if I find a significant carryover effect?
If carryover is detected — typically by testing a sequence-by-period interaction — the standard recommendation is to restrict the analysis to data from the first treatment period only, effectively treating it as a parallel-group study. This sacrifices the power advantage of the crossover design but avoids the bias introduced by differential carryover. Prevention through adequate washout is far preferable to remediation.
How is a crossover laboratory experiment different from a repeated-measures design?
Both designs measure the same participants under multiple conditions. The crossover design specifically involves two or more distinct treatment interventions administered in separate periods with a washout interval between them, and treatment order is randomized across participants. A repeated-measures design may involve multiple time points under the same condition (e.g., longitudinal follow-up) or multiple stimuli within a single session without washout. The crossover structure introduces the specific concern of carryover and period effects that require dedicated design and analytic attention.
Is blinding possible in a crossover laboratory experiment?
Yes, and it is strongly recommended when feasible. Single blinding (participant unaware of treatment order) reduces demand characteristics and expectancy effects. Double blinding (both participant and experimenter unaware) provides the strongest protection against performance and assessment bias. In pharmacological lab studies, identical-appearing capsules or infusions make double blinding straightforward; in behavioral or cognitive studies, blinding the participant to the specific condition being tested requires careful stimulus design.
Sources
- Jones, B., & Kenward, M. G. (2014). Design and Analysis of Cross-Over Trials (3rd ed.). CRC Press. ISBN: 978-1439861424
- Crossover study. Wikipedia. link ↗
How to cite this page
ScholarGate. (2026, June 3). Crossover Within-Subjects Laboratory Experiment. ScholarGate. https://scholargate.app/en/experimental-design/crossover-laboratory-experiment
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.
- AB DesignExperimental design↔ compare
- Crossover Randomized Controlled TrialExperimental design↔ compare
- Factorial Laboratory ExperimentExperimental design↔ compare
- Laboratory ExperimentExperimental design↔ compare
- Repeated-measures ANOVAStatistics↔ compare