The CBD Scan Trial Used Equal Allocation Despite Planning a Two-to-One Ratio
The scan-anxiety study ultimately randomized twenty-five participants to each treatment. The paper disclosed that its original sample-size plan anticipated a two-to-one allocation instead.
The discrepancy concerned the allocation ratio
The central registrar inadvertently implemented a one-to-one scheme rather than the originally intended unequal ratio. Participants still underwent randomized assignment, and the reported treatment groups were equal in size. This should not be described as a switch to nonrandom treatment selection.
The authors recalculated the planned comparison’s power
For the assumed effect and variability, the equal allocation increased calculated power from 80% to 89%. Those are design calculations conditional on assumptions; they do not represent the probability that a particular published finding is true. Equal groups also did not guarantee statistical significance.
The observed primary result controls the conclusion
The primary anxiety-change comparison remained nonsignificant. The disclosed allocation discrepancy therefore cannot be used either to invent a positive primary result or to declare the whole experiment nonrandomized. Reporting what was planned, what occurred and how the authors assessed its consequence provides a more precise interpretation. The secondary anxiety signal remains hypothesis-generating within a small, single-dose trial of a specified clinical population.
Related analysis: the study’s broader findings and limitations.
