Code sharing practices in biomedical research:
New study finds that approximately half of the analysed manuscripts did not share associated code.

Preprint evaluated on @prereview by @polka and and Stephen Gabrielson.

Excellent post-mortem of peer-review.io - why is it so hard to disrupt peer review?

social pressure to review
need for moderation
reputational lock-in

Would love to see more scholcomm experiments report learnings like this! http://www.theroadgoeson.com/crowdsourcing-peer-review-probably-wont-work
ht Subbiah Arunachalam