>
>I like Alan's quote very much (it is on my website -
>http://www.toontalk.com/English/adultask.htm ). My question is why after 17
>years do so few people understand that that is what is so special about
>computer programming. Why are there so few actively trying to spread the
>power and joy of programming to the wider world?  Why is it such a hard
>sell?

In my opinion, Ken, the answer to that question has got to be on the 
level of a Grand Challenge.  Why is programming so hard?  (And 
similarly: Why are so many willing to move kids towards computers as 
appliances but not as media?)

At GaTech, we recently started a bunch of assessments of our students 
and found that even 40% of our 2nd Semester students couldn't write a 
dozen lines of Java code (insert and delete of a linked list) when 
given the class definition and method headers, a compiler they knew, 
and 90 minutes.  I've mentioned these assessments on this list 
previously.

Over last Spring, a group of us did a similar study at GaTech, 
another US institution, one in the UK, and one in Poland.  The 
problem was building an RPN or infix calculator (different problems 
with different data sets).  We did our meta-analyses last month. End 
result: Average grade was 25%.  (This'll be appearing in the ACM 
SIGCSE Bulletin soon.)

What's amazing is how few people are studying this.  At a time when 
so many (e.g., the PCAST report) point to the importance of 
information technology and education about it for the health of the 
economy, research into CS Education has virtually disappeared.  The 
last Empirical Studies of Programmers workshop was cancelled due to 
lack of submissions!

I think that we have clues about why it's so hard for so many, and 
how to make it easier.  Amy Bruckman showed that making programming 
socially cool (in Moose Crossing) made it possible for lots of kids 
to learn to program. The recent American Association of University 
Women report pointed out that women (and many others, I'll bet) are 
drawn away from IT because we make CS classes as boring as possible. 
Motivation is important, and showing that computers can do 
interesting and creative things is critical to making programming 
learnable.  The Squeak tile approach of using multiple parallel 
scripts (similar to the approaches in Logo Mindstorms and StarLogo) 
to avoid explicit conditionals or iteration is a fascinating 
direction to explore.

Seymour Papert has written that it's inherent in school's structure 
to avoid revolution, and that programming will never become 
integrated into school curricula because of that (ref: Journal of the 
Learning Sciences article last year, and in the Children's Machine 
and Connected Family books).  I hope he's wrong, but I think the 
small number of people pushing on this problem makes it more likely 
that he'll be right.

Mark

--------------------------
Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280
Associate Professor - Learning Sciences & Technologies.
Collaborative Software Lab - http://coweb.cc.gatech.edu/csl/
(404) 894-5618 : Fax (404) 894-0673 : guzdial at cc.gatech.edu
http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html