Re: Solving Homework in Pairs

On Wed, May 23, 2007, Shlomi Fish wrote about "[hackers-il] Solving Homework in Pairs":
> I later compiled my reasons as presented in that email here - 
> http://xrl.us/woh3 . I'd love to hear what you think about it.

Ok, here's my retort ;-)

First, before we can discuss what's *wrong* with this "pair programming" as
you call it (I call it "hagasha bezugot" - namely the hagasha, not the
programming, is done in pairs), we need to understand what's the heck is so
good about it that you want it to stay.

According to you, the good sides of this hagahsa bezugot is that "When two
partners ... sit down to solve it there can be wonderful synergy", and that
"it's more fun than solving it all alone". I fully agree with you that if this
is the case, then hagasha-bezugot can be great. The question is, in how many
course that you took, was it like that, and in how many other courses it
was something completely different, in ways I already described (e.g., one of
you slacked, or you divided the exercises between you and each one was
responsible for one week's exercises)?

I don't think it's (just) that I'm anti-social. In fact, when I was studying
I had a friend with whom I often discussed the most difficult exercises. But
we made a point of it never to discuss an exercies before we each try to solve
it on our own, and because we admired each other's intelligence, we made sure
we only gave one another "hints", rather than entire written solutions to copy.
Perhaps we were too idealistic, perhaps we had too much times on our hands,
I don't know. But we studied the way we did, and both got very good grades,
so I guess it worked (or at least didn't hurt ;-)).

On the other hand, I took once a CS course with this same guy. We were asked
to hand in the exercies in pairs, and naturally we chose each other. We ended
up deciding that he will do one week's exercise, and I will do the next. We
never discussed any of those exercises, because we both trusted each other
to do them well. We were never worried that not doing some of the exercises
will hurt us in the test, because we both understood the material well, and
we knew that the exercises (involving handing in a computer program) had very
little to do with what a test would look like (done on paper, of course).

So, to summarize, in my experience, there is not much correlation between
hagasha bezugot (two people handing in an exercise together) and pair
programming (two people working together to solve an exercise). This makes
most of the arguments you present in favor of pair programming irrelevant to
hagasha bezugot.

During my wife's studies in HaifaU, I saw even more absurd examples of
hagasha bekvutsot. At one time she was forced (not presented an option, but
forced) to hand in a work in group of 6. Since there were no "natural" groups
so large, they had to group in ad-hoc groups. Her group basically had only 2
people doing the whole work, and 4 professional slackers. The two doing the
work (my wife and a friend) didn't bother to do anything to change the
situation, because they knew that the other 4 would only hurt their grade if
they actually contributed anything to the work ;-)

-- 
Nadav Har'El                        |     Wednesday, May 23 2007, 6 Sivan 5767
nyh-TS7m/[email protected]             |-----------------------------------------
Phone +972-523-790466, ICQ 13349191 |The knowledge that you are an idiot, is
http://nadav.harel.org.il           |what distinguishes you from one.