Read the methods section. It is the only part that says whether the conclusion is allowed to exist.

Three questions kill most of the confidence in a thread like this: sample size, blinding, who paid.

The last ten percent of a taper takes as long as the first ninety. Nobody says so until you are in it.

Split orders. Two small parcels beat one large one for reasons that have nothing to do with price.