This is one of those cases where my instinct is to believe the result because I agree with the basic conceit. Given the state of replication in social sciences

This is one of those cases where my instinct is to believe the result because I agree with the basic conceit. Given the state of replication in social sciences, though, I'm inclined to be very skeptical of the paper at first blush, given that it is a meta study [1]. Especially since many of the studies cited in the meta studied show no effect or a negative effect, and it's completely possible that additional studies showing no effect or a negative effect were never published or even terminated early because null results are uninteresting. [1] https://www.gwern.net/docs/psychology/2018-gordon.pdf https://www.gwern.net/docs/psychology/2018-gordon.pdf