> I don't want to be pessimistic, rather realistic, so let me add this to > my list above > > - Constant translation thousands of items in Japanese, Spanish, > Portuguese, German, ... and trying to do that in a open process where > many translators have to be co-ordinated. Whom has worked on > localization of bigger projects will get goose bumps when reading this ;-) > - Get some kind of corporate identity into the items. I mean, that they > must read as if they where written by the same author, so that > candidates won't have to add to their examination stress, the different > slangs (localisation vs. localization for example). You need a very very > good editor team and very specific item writing policies, to achieve > that, else you would end up re-wording the most of the items (been > there, done that :-( ) > > I'm glad that Alan reported some bits of his after all successful item > writing workshop. He had an average of about 10 items per author, that > seems bad, but is surely not! But it means that you need 700 people > writing 10 unique items each, which arises the next challenge: > > - Implement a system that would distribute the responsibility of a > specified amount of items per knowledge area of each topic, to prevent > people writing too many items for popular topics and leaving out others. > I did that task twice and it was the most difficult part of all, > although I had to handle very few authors. > > - Keep Timelines with a couple of hundreds people involved. I think I do > not have to comment this challenge at all. > > In order to be constructive, I've thought a possible way out of some of > the traps I described: > > If we would publish a big pool, but not remove any faulty items and not > mention how many of the published items are really scored in exams, then > it could work. The disadvantage of such a proceeding would be that a lot > of resources are waisted (author's time, translation costs, difficulties > with the psychometrical process, longer exam duration, because faulty > items must also be presented - or the braindumpers would find out the > "real" items very fast). > > Time to stop! Again this presents only my opinion and I'm curious to see > some good arguments against of what I wrote...
Probably a better idea would be to publish the questions only and leave out the answers. This would be useful to the candidates since in their search for answers they'll pick up practical knowledge. This will also put off any potential cheaters. The other alternative to a pool of questions is to introduce a practical exam along the theoretical. Cheers, Stoyan _______________________________________________ lpi-examdev mailing list lpi-examdev@lpi.org http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
