We gave a quiz in Dr. Dalton’s 507 DE class last night. As TA, it was my responsibility to design the quiz.
There arose a problem in that several of the questions, worth the bulk of the quiz’s points, required reference to supplementary material. When the quiz is given locally this is no problem, as the supplements can be appended to the quiz packet as appendices. In the electronic environment in which we work, however, such is not an option.
For the previous quiz I’d tried embedding a link within the quiz text to an external document containing the supplementary material. This proved disastrous, though, when it turned out that WebCT insisted on opening the supplement inside the quiz window, which window had no nav bar and so left most students unable to figure out how to get back to the test.
(Bad, bad TA for not testing that in advance. Bad!)
I figured I would lick the problem this time by sending out the supplements in Word format to everyone’s WebCT mailbox before the quiz. I made up four docs, named the files with the same phrase used to refer to those files within the text of the quiz, and shortly after class began did the mass mailing.
What I did not do was include the unique identifier for each document within the documents themselves. So use to thinking of filename as identifier for contents, it did not even occur to me to include the identifier inside the text. Bad, bad TA.
So it turns out that, at least for many of the students’ set-ups, WebCT email truncates every downloaded filename at the first non-breaking space. Instead of downloading files named “Sample Search A” and “Sample Search B” to their desktops, then, many (most? all?) of the students ended up with files named Sample(1).doc and Sample(2).doc. With no internal signifiers of content, and as they’d downloaded all the files at once instead of opening them from their email as needed, at least half the class ended up having no idea which appendices were to be consulted in responding to each question.
It ended up being, frankly, a clusterfuck.
It wasn’t bad planning that caused disaster, I think, but bad assumptions. (Which, I guess, could actually be considered a special case of “bad planning.” Huh.) I had assumed WebCT would work like, y’know, everything else and download the files with the original names intact. I had also assumed students would open files as needed instead of all at once. I had also assumed it would be safe to rely on file names as unique identifiers.
These were all bad assumptions, and they resulted in a nightmare.
So, again: Bad, bad TA!