Path: utzoo!utgpu!watmath!clyde!att!ucbvax!decwrl!purdue!bls
From: bls@cs.purdue.EDU (Brian L. Stuart)
Newsgroups: comp.edu
Subject: Re: The need for an advanced degree
Message-ID: <5802@medusa.cs.purdue.edu>
Date: 9 Jan 89 17:14:19 GMT
References: <8901041445.AA20933@decwrl.dec.com> <18730@agate.BERKELEY.EDU> <911@wilbur.unix.ETA.COM>
Reply-To: bls@cs.purdue.edu (Brian L. Stuart)
Organization: Department of Computer Science, Purdue University
Lines: 69

In article <911@wilbur.unix.ETA.COM> phinker@eta.unix.ETA.COM (Paul Hinker) writes:
>
>  I've talked to quite a few people who've gotten advanced degrees and the
>general concensus is : If you're going to get a PhD then don't fool around,
>go for it and skip the masters degree.  Is this good advice?  I do want to
>get a PhD, ultimately, so I guess it makes sense.  Comments?
>  Also, which schools offer the best advanced degree education?  I've got 
>a book that lists Graduate Engineering Schools but I'd like some other input.
>I've also been told "If you're going to do it, do it right and go to MIT or
>some other 'big name'"  I guess I'm not looking for some of the obvious 
>choices here.  I got my BS at a 'no name' engineering college and find that
>I have as good, or better, education as graduates from the 'big name' schools
>with an equivalent degree.
>
>Paul Hinker

Since several people seem interested in this topic, I'll put in my $0.02 worth.
I came back to school after three years in industry mainly because I wanted
to go into achedemia rather than stay in industry.  Therefore, I can't
comment on what a graduate degree can do for you in industry.  However,
I do have a few ideas about your grad school decisions.

On the masters vs. Ph.D. question, I tend to agree with the advice you've
already received.  If you are planning on going for a Ph.D. without stopping,
then don't bother with a masters unless:
  1)  You wan advanced degrees in two different areas.  I.E. An M.S. in
      E.E. and a Ph.D. in C.S.  (This is the way I've ended up doing it.
      Not really by plan, but that's just the way it's worked out.)
  2)  The master's requirements for your program are a proper subset
      of the Ph.D. requirements.  In other words, if you can just claim
      a master's somewhere along the line without doing any more work
      than filling out some forms, why not?  This way if something happens
      where you can't finish, you still have something to show for your
      work.

A related issue is thesis vs. non-thesis masters.  Here, I feel that if
you are going on to a Ph.D. then don't bother with a thesis.  (This is
sort of a corolary to (2) above.)  On the other hand if you are a `terminal'
master's student, then a thesis will be very benificial and in some ways
is the main point of graduate school.  A thesis can really help you learn
to focus you reserch and development efforts in industry (which is probably
where you'll be since its almost impossible to get a tenor track position
without a Ph.D.).

Whether or not you should go to a `name' school is a good question.  I think
that there are pros and cons to both sides.  Going to a school with a lesser
known name or one that is weak in your area can give you a better chance to
work on research that is really your own as opposed to working on a small
part of your major professor's existing project.  Conversly, sometimes its
good to be a small part of something big, and in `name' schools, the course
offering is generally better and you have a chance to study under famous
people.  (The last point is actually worth more than it may seem on the
surface.)  I can't really comment on any schools other than those that I've
attended.  I did my M.S. in E.E. at Notre Dame.  ND doesn't have a real
C.S. department, so its not the place to go if you want a C.S. degree, but
the E.E. department is oficially Electrcal and Computer Engineering and it's
not bad in some of the networking and operating systems areas.  Currently,
I'm working on a Ph.D. in C.S. at Purdue.  Purdue is strong in Operating
Systems, Networking, Numerical Applications (there is a big physical modeling
project going on) and there is a strong push in Software Engineering (but
then who isn't pushing in that?).  The department also has some work going
on in languages and has several strong theory poeple, but it's pretty thin
in A.I.  (That's my area..)

Hope this helps...

Brian L. Stuart
Department of Computer Sciences
Purdue University