Date: Mon, 12 Jul 1999 22:46:12 -0500 From: Jonathan Lemon <jlemon@americantv.com> To: mavery@mail.otherwhen.com Cc: advocacy@freebsd.org Subject: Re: FreeBSD Certification...again Message-ID: <19990712224611.48334@right.PCS> In-Reply-To: <199907130227.VAA23721@hostigos.otherwhen.com>; from Mike Avery on Jul 07, 1999 at 09:23:24PM -0500 References: <local.mail.freebsd-advocacy/199907130029.TAA23612@hostigos.otherwhen.com> <199907130227.VAA23721@hostigos.otherwhen.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Jul 07, 1999 at 09:23:24PM -0500, Mike Avery wrote: > > Generating and validating the test is the _hard_ part. Calculating > > the reliability and validity of the test, using factor analysis to > > weed out the useless questions, and insure that scores wind up with > > a normal distribution takes a while. > > > I asked a professional (my wife, actually, :-), and she said that > > it would take a minimum of a year to develop a reliable test. The > > curriculum is the easy part. > > Ahhhh.... yeah. It's always easy to say your part is the hard part and > the other guys is the easy part. Generating a curriculum is not easy. > If it is intended to work, challenge all the students, without putting > those at the far ends of the bell curve into a coma. I didn't mean to imply that it was trivial to develop a curriculum. However, the point of a certification _should_ be to measure knowledge and ability, and so should be logically detached from the curricula itself. In other words, it shouldn't matter how they obtained their skillset, so the testing should be general enough not to be biased towards a particular training method. > Once a test has been created, determining which questions are > discrimators is pretty easy. Determining WHAT they discriminate > can be hard. Hmm. I sometimes think that psychometric specialists and educators are at times diametrically opposed. :-) > Like I said, if we don't care about validity, it's easy. If we want it > to *MEAN SOMETHING* then it gets a lot harder. I think this is the key point. If we don't really care about validity, then it's kind of pointless, really. All you do is open up a one-week workshop, and all attendees get an automatic "certification". Isn't this what Red Hat does? This doesn't mean anything, and is just another way for a company to funnel more money into their coffers. > It's a chicken or the egg problem. Has FreeBSD reached a critical > mass where it seems to matter to employers whether or not people > are certified for FreeBSD? > > What differences would you expect to see in a FreeBSD certified > person and a Certified Unix Admin? (Sorry, can't remember the > names of the group that handles that certification right now....) I think a more pertinent question would be whether employers would value a "FreeBSD certificate" over a "Generic Unix" certificate. I would expect a 'FreeBSD' specific person to be familar with the specifis nuances of the system, and be able to start working immediately without running into "compatability" type blunders. (terminfo/termcap, lp/lpr, group wheel, rc/runlevels, that kind of stuff). Personally, I wouldn't see that much difference. However, I know several people who, once they learn a specific system, will always try to make other systems behave the same way instead of learning the behavior of the new system. For these people, having a brand-specific certification is useful. I would also like to think that sites that are running FreeBSD are more likely to be better educated as to what is required of a unix admin, and accept a generic certification instead of a more "safe" specific certification. However, for all I know, this is wishful thinking. -- JOnathan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990712224611.48334>