IMHO we should go beyond this. In biology and medicine, you need to spend a significant amount of money to catch errors, as you have to reproduce huge costly experiments. I think NIH, Wellcome and other funders should consider creating a reproducibility track where they give away grants to replicate recent key findings.
I think this would speed up research by simultaneously discouraging fraud and also by increasing credibility of legitimate findings. It would also be nice if top journals created a new category of article, Replication Study, for these big efforts to replicate a substantial finding.
For example, the initial CRISPR discovery took ages to get applied to DNA editing because initially nobody believed in that finding. We could be >5 years ahead with the right incentives.
Authors with the most problematic papers, i.e. with fabricated date, p hacked, etc., will be the least likely to cooperate (‘Oh! Can’t share the data for legal reasons. Sorry!’). ERROR would end up looking into the least likely places for any infraction. I hope they have thought of that already. But, that was not the sense I got from reading the article.
This is a terrible idea. Unless the process can clearly discriminate innocent errors from malicious ones, it will send a chill through the publishing process. We don't want to pillory authors for making an honest mistake, it would kill the open dissemination aspect of science.
An academic paper is not the final word on any subject. Papers are always a work in progress, our best understanding of the subject matter. Later work is allowed/encouraged to question, rebut and contradict findings. It's all part of the academic process. It doesn't always work this way because of politics, appeal to authority, credentialism and a whole host of other issues. The system is definitely flawed, but this "solution" would make things a whole lot worse.
Please note that this initiative does not support pillorying authors. Pointing out errors increases quality of research, and as a scientist your most important guiding principle should be attempting to find the truth. The only people who do not benefit from finding errors in their work are the ones which have wrong incentives or are aware of their mistakes/malicious errors, yet for some reason do not want to acknowledge/fix them.
Incentivizing looking for errors will not improve the quality of science. Rather individuals will use this power to take down articles they disagree with. The road to hell is paved with good intentions.
But you can look for errors in articles and contact editors even right now. It appears you either do not understand what this initiative is or how scientific process work.
I think this would speed up research by simultaneously discouraging fraud and also by increasing credibility of legitimate findings. It would also be nice if top journals created a new category of article, Replication Study, for these big efforts to replicate a substantial finding.
For example, the initial CRISPR discovery took ages to get applied to DNA editing because initially nobody believed in that finding. We could be >5 years ahead with the right incentives.