Interesting abstract from Herzenstein et al suggests that replicable studies are transparent and confident relative to non-replicable ones.
Alas for itself, the abstract is timid. They “allude to the possibility that”.
Even if true, likely temporarily so. But I’ve requested the full text.