Friday, May 27, 2011

Letting papers go

Awhile back, a colleague and I wrote a paper and submitted it to a journal. The first round of reviews came back, and one reviewer told us our work was fatally flawed.

We went through a few rounds of back-and-forth with the editor, all the while repeating that Reviewer #7* was mistaken because of such-and-such reason.

Recently my colleague and I were examining our resubmission. My colleague drew a picture to clarify something, and I stopped dead in my tracks. "Holy crap, Colleague. Reviewer #7 is right! Our entire paper is irrevocably flawed."

We went though the data, checked a few things, and sure enough - fatal flaw. I'm not sure how I missed it the first time, I guess because I was not the first author and busy doing other things when we first submitted it.

So - 30 page paper goes in the trash. Clunk!

Now you might say, "But wait! Why can't you just fix that broken part? Write a big disclaimer within a limitations section?"

I can't fix it because it's wrong. The entire concept of the paper is flawed. Even with a disclaimer it would be disingenuous to publish this at all.

So I let it go. I'm not too sad, though. We actually re-designed how we'd do things to avoid this flaw in the future, and I am sure our next paper will be super fantastic when we write it. And in any case, there are always more great ideas out there.

(*) Not the Reviewer's actual number. 


  1. Perhaps it would be useful to publish (at least in a TR) why your result was flawed. If it wasn't so obvious to you and your coauthor, perhaps publishing the flaw would help other people avoid similar problems.

  2. It is always painful to get a really negative review. Your experience reinforces that after getting over the emotional response, it is really important to think about what the reviewer said and whether they have a point or not.

  3. Thanks for doing the right thing and not publishing bad science, just to get another line on your CV. I wish more people would follow your example.

  4. A similar thing happened to me with a single-author paper on a pet topic of mine when I was early on the TT. Receiving that review was quite unpleasant but after the initial fury passed and I gave it some serious though I really had to admit that the referee had a point. This episode got me thinking a bit more deeply about certain issues with the approach; as a result, the paper ended up significantly reworked and much better overall when I submitted it anew many months later. It was published after a smooth review process and is now well cited.

  5. Hey there software engineering lady, I browsed around your blog and couldn't find any mentioning of where are you from, and where did you study for your degree. I'm rather interested in that profession and I lack useful information about learning it.