Hacker News Reader

Comment by SubiculumCode | Hacker News Reader

I am not doubting your story, but I know plenty of solid careful scientists doing honest work and being successful. One of these successful scientists self-retracted a high impact paper after he discovered that he had made a data coding mistake. It was painful, but he did the right thing of his own accord, even though it halted work on several follow-up papers that he was drafting.

View on HN

How is their story related to you knowing scientists?

I am simply providing a counter-example of academic integrity to make the point that one's personal experiences, good or bad, may not reflect what is generally true of academica/science.

I think that probably most people show integrity... but it's a problem if review processes, editorial mechanisms, and culture reward those who don't show it.

The problem is of a lack of data publishing. All data should be published; all conclusions published (preferably with the code that generated the conclusions) so corrections can be made and improved conclusions drawn easily.

> The problem is of a lack of data publishing

Agreed. If you have any recommendations for long-term public data archival they would be greatly appreciated. OSF recently instituted a 50 GB cap which rules out publishing many types of raw data, and subscription options (AWS, Dropbox, etc.) will lead to link rot when the uploading author changes jobs or retires, or the project's money runs out. Sure, publishing summary spreadsheets is a good first step, but there should be a public place for video and other large data files. IPFS was previously suggested but the data still needs to be hosted somewhere. Maybe YouTube is the best option, despite transcoding?

I have no answers. If the scientific community were cooperative enough it could perhaps come up with a shared platform, but it'd be hard to associate budgets with shared resources.

Reply on news.ycombinator.com