Re: [Phys-l] 1869 MIT Entrance Exam

[John Denker <jsd@av8n.com>]

On 10/02/2007 02:24 PM, Karim Diff asked:

> Are tests supposed to be about "real-world" situations or about whatever 
> was done in the classroom (whether it was real-world or not)?

Here's the deal:

1) School should prepare people for life in the real world (*).

2) Teachers tend (*) to teach to the test.  This can be a good 
 thing or a bad thing, depending on the test.

3) Therefore my answer is clear:  In most cases the test should 
 be representative of the real world.


(*) with minor exceptions that we need not discuss right now.


I've done a lot of recruiting and hiring.  I don't know how many
times, when interviewing a candidate who had a Masters in EE from 
an Ivy League school, I found that the guy ate breakfast on the
complex-s plane, but didn't know how to use a soldering iron,
couldn't explain why the switch I gave him had six terminals on
the bottom (instead of two), and generally made it clear that he
had never built anything in his entire life.  And the guy with
a Masters in Computer Science could talk on and on about NP-
completeness but couldn't get a real computer to do simple tasks.

Of course it wasn't all bad;  the good candidates were really,
really, good ... but usually on account of some non-classroom
experience:
 -- work-study job in a research lab
 -- part-time job in industry
 -- ham radio or other high-tech hobby
 -- et cetera


All in all, I think that unreal tests matched to unreal class
work are a big, scary problem.