Living Without the Bunnies

Postgraduates students in the department of Electronics and Electrical Engineering at the University of Glasgow are forced to submit our yearly progress reports to the the plagiarism detection website "turn it in". Leaving aside the questionable copyright and privacy issues involved in tying academic progression to an independent for-profit company, I was amused by how badly their "service" worked. I therefore created a drinking game to celebrate its ineptitude.

Text IN BOLD CAPITALS indicates the portions of my report which were "similar" to other work.

Drink a finger's width if turnitin accuses you of plagiarising...

• "NOTE 3 2.5 2 1.5 1"... because table headers are such valuable intellectual property!
• "PROGRAM FUNDED BY THE EUROPEAN COMMISION'S FRAMEWORK PROGRAMME 6"... because this kind of phrase is so unique, it's impossible that two people could think of it independently!
• "THE FUNDAMENTAL FREQUENCY of the signal. However, FOR OTHER instruments, the frequency OF THE"... yep, gotta be careful about words like "for", "other", "of", and "the"! And obviously anybody using the words "fundamental frequency" in an engineering PhD report must be up to no good!
• "AN ALGORITHM which COULD BE tweaked TO change THE `personality' OF THE"... we're still prickly about the words "to", "of", and especially "the".
• "THERE IS VERY LITTLE DIFFERENCE BETWEEN THE bottom-center AND" ... sorry, I've run out of jokes to make about simple language.

Down two fingers' width if you've plagiarised...

• "WITH A MEAN OF 40.10 AND STANDARD DEVIATION OF"... really unique writing there. Nobody else writes about means and standard deviations, after all!
• your bibliography itself is plagiarised! Yep, using exactly the same authors, article title, journal title, volume, number, and year as other people is a sure sign that you're a pirate!
• you get flagged for a using direct sentence from another paper... which was inside quotation marks... and followed by a [citation] giving the exact page number of the quote. Err, oops. Intead of "the quote", I should have said "the blatant theft of intellectual property".

Chug the glass if you've plagiarised...

• "WORK IN COMPUTER-ASSISTED COMPOSITION HAS FOCUSED ON EITHER"... hey, I just plagiarised a phrase from an international conference paper... for which I was the primary author!
• you're flagged because you've cited the same paper in your report that you'd previously cited in an international publication for which you were the primary author. (a different paper than last time)
• you've plagiarized a citation to your own international conference paper from an undergraduate student paper from Bristol which (presumably) cited you.

All examples are presented in their entirety -- there wasn't more text connected to the statistics example which I've omitted to make a joke. It really complained about "mean is blah, standard deviation is blah".

Furthermore, the bibliography "similarity" was a huge portion. The "similarity index" of my 20-page report was 10%, but when I select the "exclude bibliograph" option, that dropped to 2%.

This is quite puzzling -- a bibliographic entries should be exactly the same as other papers. I mean, given text like this:

G. Percival, Y. Wang, and G. Tzanetakis. "Effective Use of Multimedia for Computer- Assisted Musical Instrument Tutoring". In EMME '07: Proceedings of the International Workshop on Educational Multimedia and Multimedia Education, pages 67–76, New York, NY, USA, 2007. ACM Press.

there's no other way to cite that article! (BTW, this was the citation that I apparently "plagiarised" from the student in Bristol) I mean, any deviation from the above would be wrong. Ok, I suppose I could omit the page numbers, or write the conference name in a slightly different manner (like omitting the "EMME '07" part)... but get real. The best papers would cite articles in the same way, probably even using the shared databases of bibtex entries like DBLP, the Collection of Computer Science Bibliographies, or getting a bibtex entry directly from the publisher's website). What's more, their "service" clearly identified my bibliography as such -- so why complain about the similarity in citations? That just doesn't make sense.

I have nothing against computerized grading of student work -- I mean, that's half my Masters, and will be half my PhD thesis. I'm quite aware of the pressures on universities and colleges, and the overwhelming amount of student cheating that goes on. There was an outrageous amount of plagiarism going on in the course I was teaching, and it really sucked to give grades to the cheaters (I lacked solid evidence) and see hard-working (but non-cheating) students only complete two assignments out of five. And technology definitely can play a role in keeping this under control.

But that's no excuse for poorly-designed software and idiotic academic policies. If the software can't handle submissions from people with international publications, why is it being used on postgraduates at a major university?! Restrict its use to undergraduates only -- there might still be some problems with a few senior undergraduates having published a paper or two, but those would be rare. But PhD students should be expected to have a few publications before they start, and will probably have 5-10 by the time they finish. There's only a few ways to give a short history of "related work" in your area; of course you'll use some phrases that you used in your previous publications!

Oh well. It provided a certain amount of amusement, and I don't anticipate any problems when I meet with the faculty member who evaluates my technical report and the "similarity report" about it. I have to say, though, that I'm really wishing I stayed at UVic. This is just insulting.