Hi... rather than get into details of my or Derek's personal
involvement with these issues, let me try to make it more
about our general perceptions again. Probably more interesting
for the mailing list.
On Sat, 6 Dec 2003, Derek M Jones wrote:
> > I think we must go to different SE conferences or get
> >different messages from them.
> Probably been to less than a dozen in my life.
> I was referring to how industry views academic research.
I agree that a lot of academic research in general is going
for the "maybe useful in 10 years" outcome that you mentioned.
However, I have a different impression of SE research in
particular, as published at conferences like ICSE, and journals
like IEEE Transactions on SE. For instance, Gail Murphy's group
(the discussion of which started out this thread) is obviously
aiming at things of immediate relevance to industry.
> We probably have different views about what topics come under
> SE research. Some people would claim that formal methods is
> not just a branch of pure maths, but has some connection with SE
I take the above conference and journal to be canonical SE
publications. I wouldn't expect a paper on (e.g.) programming
language semantics to appear in ICSE, but I would expect to see
papers on (e.g.) testing there.
> The issue is one of practice. A strong case can be made for showing
> that undergraduates have had plenty of practice (usually a large percentage
> of their life) performing tasks related to those they are asked to perform
> in psychology studies (in some studies involving reasoning it has been
> argued that they are not representative of the general population).
> Undergraduates have not had anywhere near the amount of SE practice
> workers in industry have had (be it good or bad). Therefore it cannot be
> claimed that the results of SE studies using undergraduates will correlate
> with those performed using commercially experienced developers.
I agree entirely with what you say here. However, in the
spirit of learning to crawl before they can walk, some
researchers have realized that, with much reservation, they must
start out by studying undergrads and postgrads before studying
professionals. As long as they discuss external validity
thoroughly and don't make any of the claims you rightly
criticize above, I think this is OK.
> Academic ALERT! The publishing treadmill. Industry is
> on the quarterly financial results treadmill (or for the somewhat
> smaller companies the lunch with bank manager readmill).
I agree entirely again, and that is basically the point that
I was trying to make.
> Given the choice between:
> a) funding for a study that you know will produce 10 publishable
> papers if no industrial relevance,
> b) funding for a study that you know will produce 1 publishable
> paper of some industrial relevance.
> which would you choose?
I assume you mean studies that would take the same amount
of effort. An untenured, tenure-track faculty member at a North
American university would read the above as a choice between (a)
doing something that will pretty much guarantee that they will
be able to stay on in their chosen career, and (b) doing
something that will pretty much guarantee that they will not.
The choice is obvious.
A tenured faculty member who is motivated by a desire to
have a positive impact on software practice would certainly
choose (b). Of course in a sense it's a fantasy scenario, since
one never knows what studies will have any impact. This is why
ICSE (for example) gives out best paper awards not for this
year's papers, but for the papers from the conference 10 years
ago that have been shown to have an impact.
--Jamie Andrews.
Associate Professor
Dept. of Computer Science
Univ. of Western Ontario
London, Ont. CANADA N6A 5B7
[EMAIL PROTECTED]
http://www.csd.uwo.ca/faculty/andrews/
----------------------------------------------------------------------
PPIG Discuss List ([EMAIL PROTECTED])
Discuss admin: http://limitlessmail.net/mailman/listinfo/discuss
Announce admin: http://limitlessmail.net/mailman/listinfo/announce
PPIG Discuss archive: http://www.mail-archive.com/discuss%40ppig.org/
