[21]

At a more meta-level, they basically failed to replicate their own paper (they disproved the main hypothesis).

Which shows exactly what the big problem is: you just throw some ideas out, do some statistical analysis, and presto, you are "a scientist discovered a new thing"

[nxc18]

This is a big success. We need more scientists willing to publish dead ends and ideas that turn out to be wrong. At the very least it hints at future areas of research. A lot of great science involves testing common sense and finding it's wrong.

They could have hacked their way to a positive result, good on them for not.

[jjw1414]

As a molecular biologist whose been in the field for nearly a quarter of a century (ouch, I'm old), I can't tell you how important your comment is. Too many times I've been to conferences where a small group of researchers is informally discussing a common roadblock in a procedure or technique and one in the group says, "We found out months ago that you can't use compound Y because it inhibits X downstream". For the love of all that is good, why didn't you publish that?! You could have saved countless combined hours and materials for other labs and increased the rate of progress. We all know the answer: Reviewers and editors want to publish flashy new discoveries, not fizzled experiments. Online discussions and open access has helped the issue, but the underlying pressure on scientists to publish only positive findings and leave out the critical trial-and-error data must end as it is is essentially fudging research and does a disservice to other researchers and science as a whole, intended or not. Whew, that was cathartic. I'll fade back into the shadows now.

[fallous]

I'd like to take the effort to thank you for staying true to the pursuit of real science and trying to demand that others meet the standards. A well-tested negative result is not a failed experiment, it is one less option to pursue.

As I've been known to tell junior engineers that ask how to become expert in some field, "the easiest way is to fail in every possible way without repeating yourself. Whatever is left is correct. Note that I didn't say that's the fastest way." Gaining new knowledge is always a combination of discovering both dead ends and possible successful paths.

[fabatka]

I completely agree, it was a really big relief when I read the last 3 sentences of the abstract. I was expecting the typical "sensationalist" ending where surprise, surprise the original hypothesis of the researchers turned out to be true, unbelievable.