Metadata, formerly "Why Your Takehome Exams May Not Be Anonymous"
UPDATE: Holly Parrish writes in to add "Professors get hard copies of takehome exams." So, no metadata problems at Boalt Hall. And thank you for stopping by and reading the blog!
--
What, what, what?
That's right. Despite best efforts to use only exam numbers, your takehome exams are not anonymous. Find one of your takehome exams from last semester. On a Mac, examine the document in the Get Info box in the finder. This can be dome by clicking on the icon and pressing command-I. Once in that screen, examine the "More Info" tab. See the "Authors" field? Behold, your name.
What's going on? Microsoft Word appends "metadata" to every file. This has been the bane of e-discovery and the subject of many ethics opinions (John Steele's blog has discussed this over the months, for example, see discussion of Florida's ethics ruling. For more, search the blog for "metadata".). Included in that metadata is the author of the document, which the application assigns based on the registration of the copy of Microsoft Word. Every document you've made lists as the author whatever name the program is registered under. (a special problem if you registered your copy facetiously as "Darth Vader" or worse, registered a cracked version of the program with a name and serial number from the Internet)
[see UPDATE above. The Boalt Hall exam process does not allow professors access to digital copies, so they cannot get into the document's metadata.]
But the metadata problem has ramifications elsewhere. One group of people that does have the sophistication to exploit these traces of data are employers. While confidentiality obviously does not matter in the employment context, many employers will care if you're using pirated software (while doing research on piracy norms, I learned that around 1/3 of employers would have serious reservations about hiring someone they knew used illegal copies of software, music, or movies). The metadata gap allows them one avenue to check on your respect for copyright norms.
(In the legal practice context, the far more damaging metadata problem arises from the autosave feature in Word and Excel, as well as old track changes information. I don't know how to pry out this metadata, but the legal ethics cases all involve peeking into the documents to observe secrets that opposing counsel thought they had deleted)
Anyway, the bottom line is simply to be careful when sending around Word documents. Consider converting them to PDF files (the conversion process wipes out all metadata). [Fortunately, the Boalt Hall exam process does not require these steps.]
(I bet you didn't think you'd be reading about legal ethics or copyright law when you read the title of this post!)
--
What, what, what?
That's right. Despite best efforts to use only exam numbers, your takehome exams are not anonymous. Find one of your takehome exams from last semester. On a Mac, examine the document in the Get Info box in the finder. This can be dome by clicking on the icon and pressing command-I. Once in that screen, examine the "More Info" tab. See the "Authors" field? Behold, your name.
What's going on? Microsoft Word appends "metadata" to every file. This has been the bane of e-discovery and the subject of many ethics opinions (John Steele's blog has discussed this over the months, for example, see discussion of Florida's ethics ruling. For more, search the blog for "metadata".). Included in that metadata is the author of the document, which the application assigns based on the registration of the copy of Microsoft Word. Every document you've made lists as the author whatever name the program is registered under. (a special problem if you registered your copy facetiously as "Darth Vader" or worse, registered a cracked version of the program with a name and serial number from the Internet)
[see UPDATE above. The Boalt Hall exam process does not allow professors access to digital copies, so they cannot get into the document's metadata.]
But the metadata problem has ramifications elsewhere. One group of people that does have the sophistication to exploit these traces of data are employers. While confidentiality obviously does not matter in the employment context, many employers will care if you're using pirated software (while doing research on piracy norms, I learned that around 1/3 of employers would have serious reservations about hiring someone they knew used illegal copies of software, music, or movies). The metadata gap allows them one avenue to check on your respect for copyright norms.
(In the legal practice context, the far more damaging metadata problem arises from the autosave feature in Word and Excel, as well as old track changes information. I don't know how to pry out this metadata, but the legal ethics cases all involve peeking into the documents to observe secrets that opposing counsel thought they had deleted)
Anyway, the bottom line is simply to be careful when sending around Word documents. Consider converting them to PDF files (the conversion process wipes out all metadata). [Fortunately, the Boalt Hall exam process does not require these steps.]
(I bet you didn't think you'd be reading about legal ethics or copyright law when you read the title of this post!)
Labels: Exams, Technology Rants
posted by Tom Fletcher at 12:12 PM
5 Comments:
http://www.netflix.com/Register
Now that was a lot of hot air! I guess we're going to be getting more of the same out of TF in 2007...
This may matter to those of you who post your outlines to boalt.org or anywhere else. If you don't want others to know it was yours, strip the metadata. There are some outlines I've looked at where I wondered if the author knew his/her identity was not anonymous.
Does anyone have a link to a good metadata scrub program?
Or is there a way to keep word from putting in the metadata in the first place? It seems there should be.
I'm not sure if this is the same as the link listed above, but what I suggest is search Microsoft's site for
"Remove Hidden Data Tool" (RDH Tool)
This is a plug-in that Microsoft provides for Office which strips virtually all data from a document and puts in into a read-only format.
It's actually a useful tool to have installed anyway, because it allows distribution of documents in .doc format without preserving all the previous edits (something many authors might want to do).
Post a Comment
<< Home