Take this comment as an extra upvote. The thing is: we need to figure out a way to make accountability to become more of a thing. Peer review isn't enough (clearly). I suspect that people should create a journal that has more strict requirements than simply peer review. I think for many forms of science, one should provide an R script with the data they used to get the results. Yes, there are still ways to fabricate stuff but it'll be more transparent. Another thing is: there needs to be an incentive for reproducible findings

Unfortunately, I don't know how to go about this