The value of rejecting a rejection

· by Brian Anderson · Read in about 2 min · (272 words) ·

Tell me if this is happened to you…

You send a paper to a journal, and it’s rejected (ok, that happens to every researcher).

You read the reviewer comments, and you find yourself agreeing. They think there is unnecessary material cluttering your paper and obscuring the contribution. Your paper is too complicated.

So you make changes to the paper based on those comments and send the paper back out to a different journal.

Congrats! Journal Two offered you a revise and resubmit (yeah!).

One problem. These reviewers think your paper lacks justification. Your arguments are trivial, and do not provide adequate depth. Your paper is too shallow.

The content you are missing is the same content that you removed from the earlier version of the paper.

DOOH!

It happened to me last week. Again. I had established a guideline long ago that I wouldn’t make changes based on an earlier rejection. The variance among reviewers and editors is high. Very high. Call it the peer review error term, and it’s big. Very big.

So once again, there was value in rejecting the earlier rejection, and rolling the dice again. An n of 1 with high variance and low predictive ability makes for a very uniformative prior.

Now that said, a reason for the rejection is that the content is good, but the writing is poor. That’s something we all struggle with, and the journey to better writing is a marathon, not a sprint.

So for me, the lesson is to have faith in my content, but question my writing. Reject the rejection before starting again. I hope it sinks in this time.