The Chronicle of Higher Education
Information Technology
From the issue dated April 21, 2006

Microsoft Word's Hidden Tags Reveal Once-Anonymous Peer Reviewers

Related materials

Text: How to Keep Your Documents Anonymous

Printer
friendly
E-mail
article
Subscribe Order
reprints

The peer-review process at many academic journals is intended to be blind, meaning that authors do not know who is reviewing their work. But a little-known setting in Microsoft Word has led to the unmasking of some peer reviewers, compromising the anonymity of the process.

Keyne A. Cheshire, an assistant professor of classics at Davidson College, in North Carolina, is new to scholarly publishing. He recently discovered the problem by accident. After submitting an article to a journal in his field, he received a reviewer report by e-mail, forwarded from the journal's editor (he declined to name the journal or editor). The report, which Mr. Cheshire said included some "hefty criticism" of his article, arrived as a Microsoft Word file attached to the e-mail message.

When Mr. Cheshire opened the document, he noticed that it seemed to have been created using a British version of Word. Curious, he clicked on the document's preferences and was surprised to see a screen labeled "Summary" that listed the name of the person who had created the document — someone in his discipline whom he knew.

"I didn't want to know who my reviewer is," he said, adding that he had no idea that the reviewer's name might be embedded in the file.

He said other names were listed in the summary information as well, including the name of the journal's editor, who had apparently also opened the document at some point, and even the name of an information-technology-support official from the reviewer's university.

Easy to Find

As it turns out, Microsoft Word automatically tags every document with an author field, based on information pulled from the user's computer, when the file is first saved. It isn't hard to pull up that tag: It is visible by pulling down the "File" menu, clicking on "Properties," and selecting the "Summary" tab. It is possible to remove the tag manually — one document at a time.

Editors at some academic journals have been aware of the issue for some time and have devised guidelines telling reviewers how to keep Microsoft Word documents anonymous. One such journal is Women in German Yearbook, which gives reviewers a one-page guide called "Anonymity Instructions for Electronic Review."

"It's something that would have never even occurred to me," said Margaret R. McCarthy, who first learned about the problem when she became co-editor of the journal about a year ago and was shown the guidelines. "This is just basic knowledge that any editor should have," said Ms. McCarthy, who is an associate professor of German and Russian at Davidson.

S. Douglas Olson, who is editor of Classical Journal and a professor of classical and Near Eastern studies at the University of Minnesota-Twin Cities, said he stumbled upon the problem the hard way: After writing one of his first reviews, he got a note from the author, who should not have known his identity, alerting him that he had left his name in the document's summary field.

"Luckily they were very positive referee reports," said Mr. Olson. He immediately informed the journal's editor, who was also unaware of the possibility that reviewers' names might be hidden in Word documents. "She was just horrified," he said.

Scope Is Unclear

Many journals, including the journal to which Mr. Cheshire submitted his article, use a double-blind process — the reviewers are not supposed to know who the authors are, either. That means that hidden tags in electronic submissions could give away an author's identity as well.

It is unclear how many journals could be affected. Not all journals use a blind peer-review process. Some journals still distribute documents to reviewers and authors by postal mail. Many journals, especially in technical fields, e-mail documents as PDF files, and in many cases, converting documents to such files removes identifying tags.

Still, use of Microsoft Word among scholarly journals appears widespread.

"Probably most production of journals and the review of most journal articles is being done in Word," said John Unsworth, dean of the graduate school of library and information science at the University of Illinois at Urbana-Champaign and a founder of one of the earliest electronic journals, Postmodern Culture.

He said that when he edits articles, he creates a new file in which he incorporates the comments of several reviewers, rather than forwarding the files the reviewers created.

A spokesperson for Microsoft, Catherine B. Brooker, said she had heard no complaints from professors about the tagging. She said that document-summary tags are designed to help businesses manage the flow of documents.

"It's metadata," she said. "From a records-management perspective, you might want to hang onto it. But at the same time, Microsoft realizes that you may not want to have all this information out there publicly."

Ms. Brooker said that a forthcoming edition of Microsoft Word, part of Microsoft Office 2007, will have a new feature called Document Inspector that will help users strip unwanted identifiers from documents. "A review process is part of everyone's lives, whether you're in academia or not," she said.

The new feature, she added, is designed to "give people more power in how that review process happens."

Several academic leaders contacted by The Chronicle said they were unaware of the problem, and some were not sure whether their editorial practices allowed reviewers' names to be revealed.

"This is completely new to me," said Jana Argersinger, vice president of the Council of Literary Magazines and Presses. "It may be that none of our submitters have been savvy enough to know about it."

"I have never heard of this," said Rosemary G. Feal, executive director of the Modern Language Association. "And I myself am a journal editor." She said that the journals she works with did not send reviewer comments directly to authors as Word files.

"We're now preparing a new edition of our MLA Style Manual," she added. "We tell a million things to authors. This is something that we might need to tell them."

Journal-Editing Systems

Some journal editors say they use software systems designed to manage the review process, and that many of those systems prevent reviewers' names from slipping out.

Margaret Ann Winker, deputy editor of the Journal of the American Medical Association, said the journal uses a Web-based system to manage its review and editing process. Dr. Winker said the system, eJournalPress, automatically converts Word documents submitted by reviewers to PDF format and clears out identifying tags.

"Prior to that we copied and pasted reviews into e-mail documents," she said. "We haven't used Word documents as attachments for that very reason."

The American Chemical Society uses a software package from ScholarOne called Manuscript Central. "It's our understanding that they've totally solved this problem," said Cheryl Shanks, vice president and director of editorial office operations at ACS Publications. "We've been aware of this since we first implemented online submission."

Other journal editors who know about the potential pitfall of using electronic documents have decided to stick with postal mail.

"That's why SEL Studies in English Literature 1500-1900 only forwards to our authors a paper copy of readers' reports," said Robert L. Patten, the journal's editor and a professor of humanities at Rice University. "In spite of institutional pressures to save postage and time by using electronic communications, we've been very slow to tamper with paper and print on highly confidential parts of the publishing process."

Spreading the Word

Since his accidental discovery, Mr. Cheshire has worked to tell his colleagues about it, hoping to preserve the anonymity of the peer-review process in the electronic age.

He informed Davidson's information-technology department, for instance, which in response has issued a fact sheet on its Web site on "Removing Personal Data From Office Documents."

"I would guess that the vast majority of folks just don't know that that's there," said Mur K. Muchane, executive director of information-technology services at Davidson.

Mr. Cheshire also informed an official at the American Philological Association, David Konstan, who like many was also unaware of the issue. Mr. Konstan is chairman of the group's committee on professional matters and is a professor of classics and comparative literature at Brown University.

"I've contacted all the members of the committee to let them know that we should be thinking of this," said Mr. Konstan.

'A Bit Curious'

Meanwhile, several professors interviewed for this article worried that spreading details on how to dig out identifier tags of Word documents could lead some people who never knew of the possibility to do some sleuthing in old files they have received.

"Who would not be just a bit curious about the identity of one's reviewers, whether kind or cruel?" said Gail L. Shivel, a lecturer in English at the University of Miami and associate editor of Menckeniana, in an e-mail interview. She said that many people will soon be "opening up old reviews they have received, especially negative ones, to see if they can find out who wrote them.

"The wielder of the stinging lash might make a permanent enemy, as scholars seem to have long memories for criticism."

HOW TO KEEP YOUR DOCUMENTS ANONYMOUS

Many scholars use Microsoft Word to compose journal articles and reviews of other scholars' work, then e-mail those files to editors during the blind peer-review process. But many do not realize the software automatically includes the author's name as a tag in the document's properties.

To see if Microsoft Word is recording your name:

The author tag may be seen by pulling down the "File" menu, clicking on "Properties," and selecting the "Summary" tab.

To remove identifying information:

In Windows: Pull down the "Tools" menu, click "Options," and select the "Security" tab. Then check the box beside "Remove personal information" and save the document. It is also possible to remove the tags manually--one document at a time — by opening up the "Summary" tab as described above and erasing the identifying information.

On a Macintosh with OS/X: Pull down the "Word" menu. Click "Preferences," and then select "Security." Check the box for "Remove personal information" and click "Save." Then resave the document. The tags can also be removed manually, as in the Windows version.

 
http://chronicle.com
Section: Information Technology
Volume 52, Issue 33, Page A41