Hi, I as pleasantly surprised to see this topic of discussion flare up again, because it is one that is much on our mind. <rant> In an ideal world, exams are 100% perfect, i.e., everyone who has the knowledge and skills that the exam tests passes, and the rest fails. In such a world, a certificate can be used to predict someones abilities: the holder of the certificate is guaranteed to have the abilities, and the rest might or might not, depending on whether they ever took the exam (and failed). Unfortunately, such the world is not :-). There are limits to the knowledge and skills you can test. The source of these limits are manifold, but they include time, money, but also limits as a result of they way you test. There is only so much you can test in a multiple choice exam! So, then the exam becomes a stochastic process: some people who should fail, pass (a false positive), and some who should pass, fail (a false negative). As mathematicians can explain you, when you take measures to decrease the number of false positives, you will almost automatically increase the number of false negatives! Thus, in order to keep the number of disappointed people on this planet down (the false negatives), we have to take a number of false positives (i.e. nitwits with a certificate) for granted. But, see social and economic factors kick in! As more weight is attached to the certificate, the social and economic desirability of the certificate increases: more people will try and acquire the certificate. Mathematics tells us that this gives us a greater number of false positives (as an absolute, not as a percentage), stated differently: even more nitwits with a certificate. Selective memory then kicks in: everyone has met a nitwit with a certificate and relates of this experience whenever suitable (and often when not :-). It is easier to forget all the able people who happened to have a certificate. The mere fact that false positives exist somewhat diminishes the value of a certificate for predicting someone's skills. However, I think social factors account for a larger diminuition than is mathematically justified. However, I think another factor comes into play. When the economic/social desirability of a certificate becomes great enough, people will start wanting the certificate not for what it stands for, but for what it gives them. Their goal becomes not knowing the things and attaining the skills that are "behind" the certificate, but knowing the things and attaining the skills that allows them to pass the exam. Put simpler: they will explicitly start training to pass the exam! I know that in the MCSE domain there are exam trainings that focus on the skills needed to pass the exams: exam structure, question structure, sample exams, exam strategies and so forth. I have encountered people in an MCSE training whose sole motivation for following the training (and getting the certificate) was that it would allow their employer to boost their hourly rate with 20 guilders (+/- US$ 9,=)! I taught a couple of so called "master classes" that would lead people to the MCSE status in three months. Some of these folks were pretty bright and knew how to pass exams: they could study towards that goal, maintain most of the necessary knowledge in short-termish memory, pass the exam and then do a memory flush so that there was enough space left for the next topic of discussion. And, believe me, these are not isolated incidents!!!!!! Well, such is life, and it gets sucher every day :-) We would not want a certification program that only Linus himself can pass? Would we? Life is not perfect, let's just do our best. </rant> Jos Visser -- Broaden your education by reading: http://www.musa.org.
begin:vcard n:Visser;Jos tel;fax:+31-20-4950 223 tel;home:+31-36-522 0400 tel;work:+31-20-4950 222 x-mozilla-html:FALSE url:http://www.osp.nl/~josv org:Open Solution Providers adr:;;Dalsteindreef 16;Diemen;;NL-1112 XC;Kingdom of the Netherlands version:2.1 email;internet:[EMAIL PROTECTED] x-mozilla-cpt:;0 fn:Jos Visser end:vcard
