Reader Comments

Post a new comment on this article

"sample size calculation"

Posted by daniel_hicks on 13 Oct 2015 at 23:42 GMT

What does a "sample size calculation" mean here? The term does not appear to be defined in either the paper text or the SI. Based on the very low numbers, I assume this means something more than just reporting the sample size. The discussion of "sample size calculation" on page 4 of the SI makes it sound like it might be a pre-study power calculation. Is that right? Thanks in advance for the clarification.

No competing interests declared.

RE: "sample size calculation"

mmacleod1975 replied to daniel_hicks on 14 Oct 2015 at 07:06 GMT

Dear Daniel,

Yes, you are correct. the point being that there is a "sweet spot" for sample size, rather like the temperature of Goldilock's prrige; too big and you waste resources and animal, too small and your results are imprecise and potentially misleading (once publication bias swings into force). You find the sweet spot though a sample size calculation or a power calculation, and you reassure your readers that you have done so by showing your working. Extra-ordinary results absent a sample size calculation are at greater risk of being due to the play of chance.

Best wishes,


Competing interests declared: I'm first author of the paper in question