12 March 2018

How anonymous is “anonymous” in peer review?

Last time, I was musing about the consequences of signing or not signing reviews of journal articles. But I got wondering just how often people sabotage their own anonymity.

As journals have moved to online submission and review management systems, it’s become standard for people to be able to download Word or PDF versions of the article they are reviewing.

The last article I reviewed was something like 50 manuscript pages. There was no way I was going to write out each comment as "Page 30, paragraph 2, line 3," and make a comment. I made comments using the Word review feature. And all my comments had my initials.

As more software uses cloud storage for automatic saving features, more software packages are asking people to create accounts, and saving that identifying information along with documents. Word alerts you with your initials, but Acrobat Reader's comment balloons are little more subtle.

Ross Mounce and Mike Fowler confirmed that this happens:

Yep. Metadata tags are great. 😀 Even simply the language setting can be a giveaway: Austrian English is a huge clue in a small field [real, recent example!]. "Blind" peer review is not always effective...

Having wondered how often authors do this, I wonder if editorial staff ever check to make sure reviewers don’t accidentally out themselves.

Picture from here.

1 comment:

Antoine Souron said...

As a Handling editor of a paleontology journal, I always check the metadata of the files uploaded by the authors. Most of the time, there are information that would betray the authors' anonymity (full names or initials).
I wonder whether most editors check that before sending the files out for review...