Allocation concealment as a potentially useful aspect of randomised experiments

摘要

In experiments where subjects are allocated to different treatments, implementing allocation concealment simply means that procedures are used to prevent conscious or unconscious human bias influencing the allocation of particular subjects to particular treatments. It is a related, but distinct, procedure to blinding. Allocation concealment is a neglected, but potentially valuable, tool in improving the design of experiments, and it can always be applied inexpensively and easily to any experiment involving allocation of subjects between treatment groups. I feel allocation concealment should be adopted more widely.