On 10-07-14 12:42 PM, G. Matthew Rice wrote: > I spoke with our psychometrician and others about the idea of creating > a pool of 3,000 to 5,000 questions which were all publicly available. > With this many questions in the pool, memorising all of the answers > would not really work without (acquiring?) an understanding of how the > technology worked. > > In this case, I'd propose a pilot project like thus: > > 1. Pick an exam > 2. Attempt a community project to create, review and assess the items. > 3. Publish a couple of "open" versions of the exam along with the > usual versions. > > If things work out, we could try it with other exams. If they don't > work out, we stick with the current model.
You could keep the current set of question for one specific exam (say. 101) secret, and continue to deliver this set of question to candidate. In the meantime, open up the contribution pool and start the QA process for community-contributed questions (peer evaluation, or voting). Once you have 3000 community-contributed and approved questions, switch to those and open up the (current) secret set of questions. That would mitigate the possibility of community contribution not picking up: you would always have the possibility of sticking to the current secret set of questions. _______________________________________________ lpi-examdev mailing list lpi-examdev@lpi.org http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
