Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!lll-winken!uunet!sdrc!diluthr
From: diluthr@sdrc.UUCP (Dan Luther)
Newsgroups: comp.edu
Subject: Re: CS-1
Summary: Programming vs problem solving
Message-ID: <715@sdrc.UUCP>
Date: 8 May 89 16:08:34 GMT
References: <2130@iitmax.IIT.EDU> <2394@brahma.cs.hw.ac.uk> <206@psgdc> <2790@buengc.BU.EDU>
Organization: Structural Dynamics Research Corp., Cincinnati
Lines: 46

In article <2790@buengc.BU.EDU>, art@buengc.BU.EDU (A. R. Thompson) writes:
> In article <206@psgdc> rg@psgdc.UUCP (Dick Gill) writes:
> >[...]
> >It is important that the begining programmer understand  that
> >the first step in writing software is to structure a clear and
> >logical solution to the problem begin solved.
> 
> It has been my experience that the biggest difficulty beginners have is
> learning how to state the problem clearly and concisely.  I have found
> that once students grasp the idea of clean problem statement that the
> solutions come rather easily.

This is so true.  At Cal. State Chico where I got my degree, the initial CS 
class (read 'thinner class') was assembly language programming (two thirds of 
the initial enrollment had dropped by the final).  One of the reasons this class
was so tough was beacause, in the words of my TA 'it's really two classes in 
one, problem solving AND assembly language programming'.  He was right on the 
target!

Although I liked the class, I think the problem with tying a programming 
language to the process of learning to solve problems places artificial 
constraints on the student.  IMHO it would be more benificial to teach the 
student good problem solving skills independant of a computer language (BTW
Kepner Tregoe does a very good job of teaching structured problems solving
techniques).  I believe that once the students have developed good problem
solving skills, programming then becomes more a matter of taking advantage of
the features of a specific language to implement the solution (on a side note,
in an ideal environment part of the process would be determining *which* 
programming language was best suited to the particular *solution*)

Going back to my original reason for responding, as A.R. Thompson writes, yes
I'd say 50 to 75 percent of problem soving skill comes from proper explicit 
problem definition.

 
And to address another thread, if you are going to teach students how to solve 
problems with computers, use an 'english' style psuedo code.  Worry about the 
programming mechanics later.

	Dan Luther
	uunet!sdrc!diluthr
	(513) 576-2452 (w)

Disclaimer: I am not associated with K.T. in any way, shape, or form, other
than having had the pleasure of attending one of their classes and being very 
pleased and impressed.