Blackboard revisited

Thanks to the recently discovered Blogger Stats, I discovered that my most popular post of all time was my 3-year old rant, "Blackboard Sucks" at some 300+ pageviews, while the vast majority of my posts have nary a page hit in sight. And a leading search expression that found it was "Can Blackboard tell if you cut and paste?".  Well, going through a stack of last assignments and term papers, I can safely say that Blackboard in and of itself can't tell, though a related product, TurnItIn seem to be so capable of detecting plagiarized works, though only against stored works. 

(One important tool in detecting plagiarism is the grader's intuition. An uncharacteristically well-composed paragraph not quite relevant to the assignment topic, and/or in a different font and color from the surrounding text can be a giveaway. And after a couple of semesters teaching, the writing style of Wikipedia is jarringly different from most student writing to be suspect.  And as I remind students continually, Google works both ways, no matter how obscure the content source would seem to the typical 18-year-old.)

Where I suspect these products fall short is in a couple of places: an adequate understanding of citations and microcontent, and student copying.  First, the understanding that a document is made up of sentences and paragraphs would help the determination of what portions of the document may have been attributed or plagiarized. This has a good starter discussion of Web 2.0 in this context.  By deconstructing the document into its associated parts, even at the sentence or clause level, sequencing it, and a little link analysis could yield some interesting insights about the depth of research in addition to mere cut and pasting off Google search results.

The second is whether students grab paragraphs or even entire papers from others.  By comparing papers in pairs with other students within the class (or even across course sections), a richer understanding of student integrity can be realized.  Copying entire papers is easy enough to detect, but grabbing a paragraph here or there is a little harder to sift through, especially in a product like Blackboard that forces you to grade with blinders on.  In short, it's a socially-networked version of plagiarism that is as old as term papers and yet out of the grasp of the products I've heard about. Computationally difficult, but not impossible in smaller scales. Though if I thought it through, it might prove easier than at first glance...

But what Blackboard could use is this ability to compare projects side-by-side more readily. And by treating multiple sections of the same class as distinct courses entirely, the ability to coordinate across classes is significantly reduced. A Learning Management System (LMS) misses some real opportunities here in helping users actually compose materials, and collaborate with students using a microcontent manager that tracks authorship of individual bits of content.  Group project assessment would be far less of a nightmare as well...