Tasked based exams are much more expensive to develop. Also Pearson VUE and Prometric are not going to be able to support tasked based exams like they do multiple choice for quite a while.
I am currently developing 'hands-on' exams to replace current multiple choice exams. There are three models, in my mind. The first being a limited number of labs, ala Cisco. This requires candidates willing to pay for the rent of the lab and transportation to the lab. The second model is the RHCE class/exam type where you have someone setup systems and then proctor the exam. And the last is still fairly new and lets candidates login to virtual servers to perform the tasks. All three are resource intensive compared to the current LPI model. I am working on the later model and the problems that come with that model. How do you authenticate the ID of the test taker? What happens if the virtual server or internet connection dies? Do slow connections give the candidate more time? How do you keep the material from ending up on a gray-market site? Also Pearson VUE and Prometrics will not be able to universally support outside connections from their test centers. I think this is the way of the future but right now there are a lot of hurdles to cross. --Dave Stokes --- On Fri, 9/5/08, Tobias Crefeld <[EMAIL PROTECTED]> wrote: > From: Tobias Crefeld <[EMAIL PROTECTED]> > Subject: Re: [lpi-discuss] paper vs. computer based exams > To: [email protected] > Date: Friday, September 5, 2008, 9:14 AM > On Thu, 04 Sep 2008 23:41:48 +0300 doron > <[EMAIL PROTECTED]> wrote: > > > my question, is why we don't use the method of > tasks based exam ? > > It would be more expensive. You need more or faster > hardware and it is > more difficult to check the results automatically. > > > > for example : if the task is to build and configure > apache with 2 > > virtual hosts , we test the knowledge of installing > apache (using rpm > > for example) , using vi to change the configuration .. > the knowledge > > of apache configuration etc. > > AFAIK Cisco exams work this way. Generally I think some > practice should > be part of an exam - or one of several necessary exams to > get a > certificate. > > Automated theory tests like multiple choice tend to support > students who > are good in filling their memory with a lot of stuff for a > limited time > and forget everything after the exam to provide free space > for the next > exam. so their main problem is to do the exam ASAP after > learning. > Other skills like solution strategy or logical combination > are > difficult to test by automated Q&A-tests because those > can't deal with > descriptions as answer or different ways to the same > result. > > Usually you have info- and man-pages, documented config > files, howtos > and internet available where you can look for parameters > and correct > spellings. In this environment the limiting factor is time > and the goal > is a correct solution without "guessing" because > "guessing" needs more > time. In such practice tests you can't get a fast (!) > solution without > already knowing most of the more important commands but it > would be up > to you which one you already know. I believe that this is > more close to > "the real world" because if system XY isn't > running any more no one > asks you whether you look in man-pages or not as long as > the system is > on-air ASAP. > > > It's just an idea: Maybe in future exams for higher > certification > levels with some practice part will be developed. In the > past Cisco > went this way: "Beginners" tests (administrator > level) with multiple > choice and expert level tests in the laboratory. > > -- > Gruß, > Tobias. > _______________________________________________ > lpi-discuss mailing list > [email protected] > http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-discuss _______________________________________________ lpi-discuss mailing list [email protected] http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-discuss
