Not so anonymous peer review

by Henry on April 8, 2006

Fun story in the Chronicle this week, about the perennial academic pastime of trying to figure out the identity of the anonymous referee who dinged your article. Word documents preserve a lot of metadata, including, very often, the author’s name – so that if you submit your review via a Word email attachment (as many journals ask you to these days), and the journal forwards the review unchanged to the article’s author, he or she can figure out who you are without having to play the usual guessing game. I’ve been aware of this for a couple of years (I carefully strip all data before sending reviews out, just in case) – but I suspect that many academics aren’t (some of them may not even realize that Word collates this data automatically).

{ 28 comments }

1

Zack 04.08.06 at 11:12 am

Writing academic papers in Word? What has the world come to?

2

dearieme 04.08.06 at 11:19 am

But I practically never open a new Word document, I just “save as” a copy of something to hand and then edit it into a suitable denunciation. Does that mean that all sorts of innocent parties who originated the now-edited document will cop the blame? Blimey!

3

almostinfamous 04.08.06 at 11:23 am

you don’t necessarily have to write them in word… unless you’re sadistic

4

djw 04.08.06 at 11:39 am

I honestly don’t know what else I’d write them in.

5

Gar Lipow 04.08.06 at 11:54 am

One tip. You can open a word document in a plain text editor, put your keyboard in overwrite mode and replace any identifying info with spaces. It may just be my particular version, but so far this has not rendered any word documents unopenable in word. Naturally, if you try this trick, save under a new name, so you have the old version if it does not work as well with your version/settings and renders the document unopenable.

6

Mike Otsuka 04.08.06 at 12:53 pm

I think it’s easier just to go to “Options” on the “Tools” menu, then go to “Security” and check “Remove personal information from file properties on save”, and finally click “Okay”. (Or don’t, if you want your referees to know who you are.)

7

Hektor Bim 04.08.06 at 12:54 pm

LaTeX, obviously.

8

Steve 04.08.06 at 1:01 pm

This feature of Word is pesky for academics in other respects as well.

For example, simply removing the title page from an electronically submitted Word document before forwarding it–for “blind review”–to journal referees or conference program commitee members won’t ensure that those folks will be blind in the relevant sense.

9

c 04.08.06 at 2:23 pm

Word’s ‘track changes’ feature is another transmitter of unintended data.

10

abb1 04.08.06 at 3:08 pm

Notepad!

11

Eszter 04.08.06 at 3:13 pm

To generalize to all your Word documents what Mike O. describes above, follow the instructions on the bottom of this page. I did this a couple of years ago and haven’t had to worry about anonymizing my Word documents since.

12

John Quiggin 04.08.06 at 3:50 pm

Most of the journals I deal with want PDF, which gets rid of all this metadata, I think.

I only use Word in the case of incorrigible co-authors, and not even then if I get to the document before they add too many tables and similar.

13

abb1 04.08.06 at 4:12 pm

Remember this PDF story?

14

etat 04.08.06 at 4:40 pm

Slightly more creative than dearieme’s approach is to use an edited copy of a former student’s document for such things. I’ve collected a variety of interesting document ‘properties’ this way, some of which would make for anxious head-scratching were the recipient to presume they were bona-fide credentials. A very entertaining red herring.

15

Evan 04.08.06 at 7:01 pm

This happened to me — the editor asked me to use the comment feature in Word to indicate my comments (so as to save typing “on page 14 in paragraph three …” or variants lotsa times) — so I had to change the document properties so my own name wouldn’t appear. Voila, there was the author’s name.

16

Rich Crew 04.08.06 at 8:16 pm

Yes, LaTeX, obviously…

I tend to submit reviews in raw Postscript; I don’t think this gives away a whole lot, but maybe I’m being innocent…

Maybe someday the world will wake up from the there is only Word virus.

17

Barry Freed 04.08.06 at 8:48 pm

Just get a Mac already.

18

djw 04.09.06 at 12:39 am

I hate macs. Never heard of LaTeX. I didn’t know I was supposed to be unhappy with word.

19

Kenny Easwaran 04.09.06 at 3:01 am

I thought I decent amount of academics used WordPerfect instead of Word. But in my first year of graduate school I started using LaTeX for math problem sets, and then tried it out for a philosophical paper, and have never looked back. I’m glad that the beamer standard seems to be holding its own quite well against Powerpoint, at least within a suitably techie part of philosophy.

20

Sherman Dorn 04.09.06 at 7:44 am

I just save such files in rich-text format, which leaves no identifying-author metadata (I think).

21

harry b 04.09.06 at 9:40 am

I like WordPerfect much more than Word, but switched because of a recalcitrant collaborator (in the UK). Shows that I have less power in such relationships than quiggin.

22

Mark Eli Kalderon 04.09.06 at 10:20 am

Deeply puzzled that people are submitting editable files to journals. Sure, when they accept the paper they will need it in some format that they can convert to whatever they are typesetting in, but this is not necessary for the review process and Word is simply not a portable format. PDFs have embedded metadata to be sure, but depending on how you are generating them, this is easily controlled.

Let me add another vote fo dumping Word. I have been writing philosophy in a text editor for a couple of years and it is much more productive writing environment.

23

Jeremy 04.09.06 at 10:57 am

Wrong, Sherman Dorn (re saving in RTF).

Harry B: agree with you about Wordperfect. I still use WP 6.2 for DOS when I’m in a hurry: all keyboard commands = speed, flexibility, and its dictionary/thesaurus is excellent. And, unless there is a serious compatibilty requirement, I prefer (Windows) Wordperfect (currently v12) to working in Word. There is nothing to rival WP’s “reveal codes”, a feature I find it hard to do without, especially when formatting is an issue.

24

Chris Bertram 04.09.06 at 12:26 pm

Kieran put me on to LaTeX about two years ago and I’ve not looked back. There’s a lot more control, the end product looks great (so long as you use something other than the default font) and Bibtex is just fantastic for citations and bibliography. I use WinEdt as my editor, which is much less hardcore than emacs, but there are some nice features and it is really easy to use. I’ve also just started using a Revision Control System, which integrates with WinEdt and keeps track of all your changes without you having to save zillions of files called paperversion3a2.doc etc.

25

Tom Scudder 04.09.06 at 4:04 pm

The first official report of the UN’s investigation into Lebanese PM Rafik Hariri’s death was sent out as a Word document. With Track Changes data right out there. Oops.

26

y81 04.09.06 at 8:16 pm

This metadata is, as you might imagine, an even more serious problem for lawyers. There are various programs available that strip metadata from Word documents, but I don’t know what the best approach is for an individual professor.

27

cbd 04.11.06 at 11:18 am

OpenOffice allows you to control the metadata embedded in files. You can make PDFs with one-click. And its implementation of style sheets is super.

28

ingrid 04.11.06 at 1:34 pm

I’ve often been able to figure out in 10 seconds who my referees were (many referees/editors do not even seem to realise that there is such a thing as “file properties”).

Since I’m one of those sad scholars still writing in Word, I am always sending my referee report in the body of an e-mail message, without signature. Does that protect my anonymity?

Comments on this entry are closed.