Should statistically significant but clinically meaningless results be reported as significant?


Instead of calling the particular paper that led to this blog post (I also don't want to add to its Altmetrics), just a question.

If your conclusions from the systematic review prove that a particular supplement / diet / diet resulted in an average weight loss of 0.7 pounds, it is advisable to describe this result as significant even if you believe statistically that you are capable of doing so. to this claim?

Personally, I don't think so.

Especially not when discussing food, because, as Kevin Klatt recently pointed out in his blog, there is no food placement. and as Ioannis Ionidis pointed out, we eat thousands of chemicals in millions of different everyday combinations, which strongly challenges our ability to believe definitively about the impact of any food.

Even worse is the fact that the media (both traditional and social) will not be enthusiastic about describing these findings and will instead refer to them as beneficial, important and important, such as, of course , the PubMed warriors.

How to fix this? Maybe a qualifier, "but not likely to be of clinical significance" the abstract statement can lead to more balanced media coverage (or less media coverage), which in turn is less likely to report significant but clinically useless results, which would ultimately be good for science and scientific education.

.