Use when
- Subjects, cages, samples, sessions, or runs are assigned to groups or order.
- Known nuisance variables can be balanced through blocking or stratification.
- Allocation bias would change endpoint interpretation.
Study Designs
Study design
Randomization
Randomization assigns experimental units to groups or conditions using a planned rule that reduces allocation bias.
Decision summary
Use randomization when assignment could otherwise follow cage, litter, operator, order, room, or availability patterns. Define the randomization unit before assignment, and report the method clearly enough to audit.
| Allocation unit | Subject, cage, sample, run, plate, or session assigned by the random rule. |
|---|---|
| Method | Random sequence, block randomization, stratification, or other planned assignment. |
| Concealment | How assignment was kept from influencing enrollment or handling. |
| Audit field | Seed, list, block, strata, and assignment date when appropriate. |
Do not use when
- The allocation unit is unclear or differs from the analysis unit.
- Randomization is claimed after convenience assignment has already occurred.
Caveats
- Randomizing subjects does not fix cage, litter, batch, or operator clustering if those remain confounded.
- Unequal allocation can still occur and should be reported.
- Randomization and blinding solve different bias problems and both may be needed.
Reporting checklist
- State the allocation unit.
- Describe the randomization method.
- Report blocking or stratification variables.
- State who generated and applied the allocation sequence.
- Report deviations, exclusions, and imbalance after assignment.
References
Festing and Altman 2002. Guidelines for the design and statistical analysis of experiments using laboratory animals.
Study-design guidance including randomization and analysis-unit concerns.
https://pubmed.ncbi.nlm.nih.gov/12100861/
ARRIVE 2.0 guidelines 2020.
Reporting expectations for allocation and design transparency.
https://pubmed.ncbi.nlm.nih.gov/32663219/
Related surfaces
Use these related surfaces to move from the scientific method question to the relevant product page, endpoint definition, analysis tool, or adjacent guide.
