In my observation regarding getting jobs and such, I think that
generation and environment matter.
Every new faculty member we've hired in the past eight years or so has
been primarily a Stata user (apart from a couple of qualitative people).
Even older faculty have somewhat come over, if only because once
students get past their intro classes, it's all Stata or specialized
packages, so they need to get with the program to be able to work with
high-quality grad students and take advantage of the opportunities
provided by younger faculty. The person who teaches intro is great, but
won't move away from SPSS, given ten or fifteen years of notes and such.
Too much inertia. Once the person teaching intro hands it off to
somebody else, then it will be all Stata. These are my observations
regarding Sociology.
I may go over my ICPSR requests re: format, but I *think* Stata is
becoming a more popular format than SPSS in Poli Sci and other fields.
Nobody asks for SAS, because if you use SAS, you don't need your ICPSR
rep to get it into a stats package -- you already know what you're
doing. SAS is so user un-friendly that it's insane. With serious
multi-user applications the security features of SAS become relevant,
and that at its guts it's moving to SQL, SAS may have a future in an
enterprise environment. But for a beginner and/or everyday usage, not
the way to go. If you're doing hardcore multi-user stuff, it has
something to offer.
Re: jobs, given the paragraph above, Stata won't make it into the
enterprise right away. Sure, it now has ODBC support, but not to the
extent that SAS does. SPSS has it, too, but so much of it is those
silly "wizards" that make your job more difficult, not easier. So for
enterprise stuff, SAS is marketable for the foreseeable future
(especially given that it's so hard to learn). But for academia, Stata
is the wave of the future (unless you're so hard-core that you become an
R zealot, and that's too much for your average undergrad).
SPSS is for MBAs who don't understand how to properly use Excel -- 'nuff
said.
Below is the text of an off-list response I got, that describes it from
a student's perspective (who is at a different department, and I've
never met. Dunno if s/he was merely being humble, and I apologize to
him/her if sharing the words below are a breach of confidence. But it
summarizes a lot.
Ben
From someone who learned SPSS as a first year masters student and then
switched to Stata in the first year of my doctoral, I explain it this
way to students: Stata is as simple as SPSS and as powerful(and
sometimes more so) than SAS. I didn't even need a formal course to learn
Stata. Every student I know who started with SPSS and switched to Stata,
self learned within a semester and would NEVER switch back - ever....
It's kind of like going from particle board to mahogany, or from
linoleum to marble, or plastic to leather. They just have to trust you
on this!
*==========================================;
*Ben Earnhart
*Computer Consultant and
*ICPSR Representative
*Department of Sociology and
*College of Liberal Arts
*University of Iowa
*(319) 335-2887
*[email protected]
*==========================================;
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> Clive Nicholas
> Sent: Friday, October 13, 2006 9:08 PM
> To: [email protected]
> Subject: Re: st: Stata vs SPSS
>
> Joseph Coveney replied to Raphael Fraser:
>
> > Let them discover on their own. They choose the package,
> and you choose
> > the problem sets. Stata is easy enough that if they decide
> to switch
> > midway through, they won't be long in catching up on their own.
> >
> > In all fairness to the students, who will be looking for
> work soon enough,
> > take a look at the job postings page on the UCLA ATA Stat Computing
> > website (URL at the bottom of this posting):
> >
> > Posting #1 "expected to have experience applying a wide variety of
> > multivariate statistical methods to health or substance
> abuse research
> > data, managing large databases, and using statistical
> software including
> > SAS and SPSS "
> >
> > Posting #2 "At least one year of SAS programming/analysis
> experience. "
> >
> > Posting #3 "Good working knowledge of SAS and/or SPSS required. "
> >
> > Posting #4 "Extensive experience with statistical
> programming (such as SAS
> > and Stata [finally!]). Knowledge with SUDAAN and SPSS."
>
> I second this vociferously: this is exactly how it is right now in the
> commercial-sector UK job market: it's knowledge of SPSS, SAS
> or MS Excel
> that counts; knowledge of anything else counts for very
> little if anything
> at all. I've yet to come across a job advert that asks for
> knowledge of
> Stata outside of the university sector, and that's very
> depressing given
> Stata's all-round strengths.
>
> Perhaps Stata isn't marketed as aggressively towards business
> users as it
> ought to be? Should it be? Well, that's what SPSS has been
> doing in recent
> years, and they've been very successful, but it's coming at
> the cost of
> making SPSS _much_ less comprehensive in its suite of
> commands than Stata,
> LIMDEP and one or two other packages that offer a much
> broader range of
> commands at very competitive prices.
>
> CLIVE NICHOLAS |t: 0(044)7903 397793
> Politics |e: [email protected]
> Newcastle University |http://www.ncl.ac.uk/geps
>
> Whereever you go and whatever you do, just remember this. No
> matter how
> many like you, admire you, love you or adore you, the number of people
> turning up to your funeral will be largely determined by local weather
> conditions.
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
