Small study, probably not preregistered (didn't find anything about it in the description). Likely Publication Bias or p-hacking. I'm starting to get interested when it's been independently replicated.
At least a practice to include that in the abstract