CB> Yuck--I hope they drop the idea of an exit test--won't work.
MS>Exit tests could build public and employer confidence in public 
MS>education by convincing the public and employers that a kid who 
MS>got a HS degree actually _retained_ some of what he once learned.
CB>So, why not drop tests and develop portfolios, projects, etc as exit 
CB>requirements.
Ok so far as it goes (why not both), but there are some
problems with this: Who's going to store all this stuff, or
do you envision passing it along to the kid?  How do the
colleges and employers get to SEE this stuff?  Would they
even bother?  You still have the problem of subjective
evaluation between teachers, schools, etc...
CB>Example:  Most tests of this sort are multiple-guess, 
CB>most have four choices.  I can take any multiple-choice 
CB>test in the world, randomize my answers and score close to 
CB>25%.  I have done this in a class and have since done this 
CB>with my own students to prove to them it was possible.
Yes, and there doesn't HAVE to be only 4 answers; what's
more, some tests have a built-in guess factor which accounts
for this type of thing....
CB>Now, if the person taking the test is halfway literate, the 
CB>average can be raised to between 50-75% by tossing out the 
CB>two choices that typically don't make sense and chosing 
CB>between the remaining two.
You're assuming the test is poorly designed.. There should
not be 2 answers which don't make sense; on the contrary,
a well-designed test would have all choices correspond to
possible results of incorrect reasoning or calculation.  I
make it a point to anticipate such errors, and I make sure
those incorrect answers appear... Who says there can't be
free response or essays; a practical, etc?
___
 * MR/2 2.26 * For a good time, call 1-800-3IBMOS2
--- Maximus/2 3.01
---------------