It seems obvious that toxicity needs to be rooted out of open source communities, and any projects that don't do so or ignore the issue will fail to keep their contributors. But it's nice to have some real studies on this with some objective results (even if not strictly for open source software project contributions).