Was it Mark Guzdial who wrote:
	>>What would you think of a manufacturing process that threw away 
	>>10-30% of its input raw material?

I'd think it was like gold production, only much much more efficient.

People can't know what a subject is like until they try it.
If computing people think that everyone who tries computing OUGHT to say
"that's for me!".

I offer another analogy:  if 10-30% of the people who attended a football
match for the first time decided they preferred baseball, would that mean
there was something terrible about football?

People arrive at University with no idea of what it is like to _do_
mathematics or software engineering.  People arrive at a teaching
hospital with no realistic idea of what it is like to _do_ nursing.
(Of course the really bad thing is that they can leave University with
a degree, _still_ not knowing what it's really like to do mathematics
or software engineering or whatever.)

It's sad when people have to spend time and money finding out that a
discipline is not what they thought it was.  But that's why fees here
and in Australia don't kick in for about a month, so that people can
pull out of papers early without losing anything but their time.

Before we start beating our breasts about a 10-30% dropout figure,
how does that compare with nursing, or carpentry, or deep-sea fishing,
or mathematics, or media studies, or fine arts, or Law?