Re[2]: A Test to Select Competency

[doug montalbano <doug_montalbano -at- CC -dot- CHIRON -dot- COM>]

     John Gear <catalyst -at- PACIFIER -dot- COM> makes a cogent argument for

<deletia> ...some form of competency testing, at least as far as
establishing some minimum, threshold level of "fit" between the person and
the job definition.  Obviously a poorly thought-out test will result in an
unreliable ("unfair" in some sense) result.  But is the current "dating
model" hiring process so superior that it merits defending?

     <astute observation about carefully-culled portfolios>
..  But a portfolio of work done in one set of circumstances may have
little or nothing to do with how one may perform in another, different set
of circumstances.

     <other good stuff deleted, including some things I *DON'T* agree
     with but find well-argued nonetheless>

     John's reply inspires me to make this point:

     At a previous job (a scientific regulatory agency) I was charged
     with screening potential candidates for editor.  Sure, I looked at
     portfolios; but these showed me end products.  I wanted to know
     *how* this person worked.  How can you do that without hiring her?
     That's right, you test her.  The trick is coming up with a test
     that shows how a person works.  Contrary to what someone else said
     about human resources tests having consistent right answers, these
     tests should have no one right answer.  You're trying to determine
     a style of working.

     BTW, what I looked for most on the editing test was how the editor
     -- not necessarily familiar with the highly scientific subject
     matter -- handled asking questions of the SME, and how the editor
     organized the text as a whole.

     doug_montalbano -at- cc -dot- chiron -dot- com