Re: Testing job applicants

[Matt Curtin <cmcurtin@interhack.net>]

>>>>> On Thu, 02 Sep 1999 07:12:10 GMT,
    Bryan Villados <news001@macgeek.org> said:

Bryan> When I went to college to get my bachelor in Computer Science,
Bryan> the students there who learned how to code literally memorized
Bryan> complete routines and kept libraries of routines written by
Bryan> their professors. They couldn't re-write them from scratch
Bryan> simply because they didn't know how to think for
Bryan> themselves. The school I went to were at fault for this dilemma
Bryan> because the professors were too lazy to grade each student's
Bryan> code for individuality, thus they told the students to code
Bryan> using their methods.

First of all, I wish to make it clear that I find the number of people
pursuing computer science, electrical engineering, and related degrees
because "that's where the good jobs are" as opposed to because "that's
where the interesting work is" nothing short of distressing.  As
observed, this results in classrooms full of students who memorize
things rather than think.

In the spring of this year, I taught a Lisp programming course at Ohio
State.  This was only one credit hour (that's not my fault, if I had
my way it would be more like 10 :-) so the grading load was
comparatively light.  I started out with 28 students (out of a
potential 40).  Over the course of the quarter, five labs were due.
None were extremely difficult, but focused on different areas of the
language.  Some were to get comfortable with the Lisp environment,
others were to get students to understand how to look at problems
functionally, and still others were for the purpose of working with
the language and understanding some of its features.

Although I was only in class one hour per week, I spent well over 10
times that on the course.  It was difficult to do my grading and get
it back to the students in a very timely manner.  The reason the
workload was so huge is simple: I have the complaints about grading
and made it a point to understand each problem submitted, trying to
see the approach that was used, and instead of deducting points for
mistakes, I returned comments that hinted at better approaches,
provided references to other useful information, etc.  Then I had to
grade the lab again when it was resubmitted.

I did everything I could imagine to make the work interesting and fun.
I spent as much time as it took to grade and to answer students'
questions, without regard to how much time for which I was being
compensated.  But I refused to accept things that did not demonstrate
understanding of the problem at hand.  Of the 28 who started, 11
passed my class.  (Most dropped; a few failed.)  I got very good
reviews from the students who were still around by the last week.  I
wish I could find a way to ask those who dropped why they did.

My Real Job is to be the site software guru and help in system
administration.  So even though I had to make the course work in
addition to my normal "40 hours" (read: "day and night" :-) job, I
think that I probably had an easier time of devoting that sort of time
than a professor who has three or four classes like that and THEN has
to worry about stuff like getting research grants, conducing research,
etc., all of which tends to have more firm deadlines than what I'm
doing.  (Missing paper submission deadlines is something of a hobby of
mine, but my impetus to publish is because I really want to do it.
Someone working for tenure or a promotion has the issue of "number of
publications" to consider, which is something of a problem, so a
professor really can't blow off deadlines the way I do.)

On the other hand, maybe if I were a professor, I would have had a
grad student available to grade for me.  But there's still a
significant time investment there, and the grad student might not have
enough hours "on the clock" to give that sort of attention to grading.

I'm not sure what the solution is.  As long as people can make huge
amounts of money in computing, there are going to be plenty of
boneheads in line to get a piece of the action, and we're going to be
finding the workforce with an overabundant supply of morons who can't
think for themselves but know only how to regurgitate that which
someone else has told them.

Maybe if we'd present problems which require independent thinking,
make the time grade work on originality, and make the basic education
in computer science more about finding innovative solutions, we'd find
that the only people who bother to stick with it are those who want to
pursue careers in computing because "that's where the interesting work
is."  But we'd be dealing with a lot of unhappy people and getting a
lot of complaints about "unfair" the education system is...

-- 
Matt Curtin cmcurtin@interhack.net http://www.interhack.net/people/cmcurtin/