Blocked Randomization with Randomly Selected Block Sizes
Author Information
Author(s): Jimmy Efird
Primary Institution: Center for Health Disparities Research and the Department of Public Health, Brody School of Medicine, East Carolina University
Hypothesis
How can blocked randomization reduce selection bias in clinical trials?
Conclusion
Blocked randomization with random block sizes can help maintain balance in treatment assignment and reduce selection bias in clinical trials.
Supporting Evidence
- Blocked randomization helps achieve balance in treatment groups.
- Using random block sizes can reduce the risk of selection bias.
- Smaller block sizes may lead to more balanced groups but increase predictability.
Takeaway
This study shows that using random block sizes when assigning people to different treatment groups can help make sure the groups are fair and not biased.
Methodology
Participants are randomized within blocks to ensure equal assignment to treatment groups, using a computer algorithm for randomization.
Potential Biases
Selection bias may occur if the treatment assignment is predictable.
Limitations
The predictability of allocation may lead to selection bias if the block sizes are not random or if the study is unmasked.
Participant Demographics
Participants were planned to be enrolled from multiple study sites, with 50 participants per site.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website