Teaching students, Google, Wikipedia etc.

[CPL.Luke]

I think the type of plagiarism I mentioned is really undetectable at the teachin level and can only be noticed by the students.

 

However you mentioned before that you suspected a number of papers had cut and paste sections from wikipedia by the general style. Is it possible that some of the authors edit wikipedia frequently and as such the style is starting to blend into there own writing? or that they are influenced by reading wikipedia so much?

[Pangloss]

You know what I found to be the most interesting aspect of using the plagiarism-detection software? The fact that it ultimately turned into an exercise in determining the degree of cheating that the students were undertaking.

 

Whether this was indicative of a limitation in the software or an indictment of the current state of affairs is something I'm still trying to sort out. Essentially what happened is that the system gave me a report on each student's paper, showing the percentage of quotations they used compared with the overall paper size. There was a button which caused to recalculate the percentage once "proper attribution" was taken into account (which I presume is optional because there are different ways to attribute quotes, so the software may not make this second calculation correctly).

 

In the end the percentages were kinda useless. What was more effective was the side-by-side comparison screen that threw up the student's paper against the ostensibly plagiarized source that the system found. Then I could eyeball them and decide for myself.

 

But as I say, the overall eye-opener for me was ultimately the fact that there was no clear way to differentiate the "cheaters" from the "non-cheaters" at a glance. Some of those with high percentages turned out not to be cheating, while some with low percentages were more suspect, at least to me.

 

At one point I actually found myself tempted to give the students letter grades based on the percentage of misquoted material in their papers! Oh, you only had 24% of mis-attributed material in your paper? Why, you get a B+! Congratulations!

 

What a crock.

[Glider]

I do see your point, but plagiarism detection software is subject to the same weaknesses as say, spell checker software. It doesn't get context. For example, to spell shecker, "Dear Sir or Madam" is no different to "Dead Sod or Moron". Both are correct, but one will go down less well with a potential employer.

 

Plagiarism detection software is badly named as it does not actually detect plagiarism (same as 'lie detectors' do not detect lies) it simply detects identical text. The flagging system is simply based upon the amount of text recognised in other sources. For this reason, our plagiarism software always flags up all coursework at around 24% identical with other work submitted on the same module, but this is because when doing the same piece of coursework, most (honest) students will refer to the same original sources, so their reference sections will be more or less identical.

 

I agree the best use of anti plagiarism software is the side-by-side checking facility. Really, that's all the software is for. It makes checking course work against possible sources a lot faster and easier than having to search for them youself, print them out and check hard copies.

 

The only effective plagiarism detection mechanism is the tutor. The software only provides a statistical indication of the more obvious at risk papers and the original sources so that the tutor can check for themselves.

 

Software can't perform the same function as a tutor. For example, I once got a student on plagiarism that the software failed to flag. I got her because it was her third piece of coursework in Research Methods, and I had come to know how she writes, which is what told me that she had not written that piece. Software can't match things like syntax, rhythm, cadence and overall 'flavour' of writing style to an individual.

 

Basically, I think that to get sucked in by the name 'Plagiarism Detection Software' (like 'Lie Detector') is a mistake that could lead to over reliance on the software through the expectation that it can do what its name implies (which it can't). Students do the same thing with spell checker, but spell checker will never replace proofreading.

[CPL.Luke]

I think the million monkeys theorem comes into play here, if you consider all of the highschool students in the world who write essays on "the causes of WW2" you probably have all the paragraphs for every 5 page paper that could ever exist on the subject already written. Software that looked for plagiarism in a database that encompassed all of these papers would surely indicate that there is simply a world wide paper trade, and that students find it easier t track down papers written 3000 miles away than to do research.

[FreeThinker]

I am currently in the second year of my Bachelors degree. Every semester its the usual pattern; things start of great and you have so much will to search the journals, towards the middle of the semester things start piling up and you are forced to look for shortcuts (E.g. Wiki) and just before exam time its just crazy and time is of the essence (back to the net)! I do think that the amount of assessment greatly reduces ones ability to learn or, more importantly, enjoy learning! I often find myself at Uni day-dreaming about getting a few free days to read some science books. The most frustrating things is that the subject content is fascinating, but the assessment and the expectations can be overwhelming.

[scifi]

i'm totally the wikipediaholic n it gives the click away answers that to in a way extent to a convincing manner

[CharonY]

Uhmm.. quod erat demonstrandum?

[Glider]

i'm totally the wikipediaholic n it gives the click away answers that to in a way extent to a convincing manner

But if you present those answers in the style you present that post, they still won't be understood.

[timo]

... that's where copy&paste comes in.