I second the motion to learn many packages.
In academic work, I worked with concurrent active files in SPSS, Stata and excel. I converted databases using "stat-transfer" and "dbmscopy." I found that SPSS worked best for exploratory questions and intuitive issues, as well as providing presentation-grade outputs. Stata was my favorite for the stats work. In both I used file-based execution functins to ensure repeatability of runs and maintain records of the research progress. I found excel to work best for pretty graphs.
In government, I have seen R become popular because it can be employed as a standard without cost barriers.
Laura B. Nielsen, Ph.D.
Adjunct Professor of Public Policy
Georgetown Public Policy Institute
3520 Prospect Street, Suite 308F
Washington, DC 20007
WEB http://www.georgetown.edu/faculty/lbn4/
Program Analyst
Environmental Analysis Division, MC 2842T
Office of Information Analysis and Access
Office of Environmental Information
U.S. Environmental Protection Agency
(202) 566-0621
----- Original Message -----
From: Jim Battista <[log in to unmask]>
Date: Wednesday, October 4, 2006 10:39 pm
Subject: Re: [POLMETH] R vs. Stata vs. SPSS
> Michael Plenty wrote:
>
> > My company uses SPSS, as most undergraduate programs and companies,
> > however, now that I'm beginning the thinking process for graduate
> school, most
> > schools, in particular Northwestern, UChicago, and Yale, all
> recommend I
> > become familiar with Stata and R.
> >
> > Are there any major differences between the three programs?
>
> Yes. R is object-oriented, which has its good and bad points. It's
> annoying if you just want regression output, because you have to poke
> the output object to make it squeak. It's very fabulous if you
> want to
> do even basic programming, because you can grab a piece of one object
> and feed it to a custom function you've written, which will then
> outputanother object. Woo!
>
> I haven't touched SPSS in donkey's years, but my *sense* is (and by
> reputation) it's not as extensible as R or Stata. However, this could
> also be completely false. But I'll go ahead and say it in the hopes
> that if I'm wrong, some good geek will be so incensed by my
> statement of
> a falsehood that (s)he will correct me.
>
> > If so, what? and what resources are out there to help me adjust?
>
> There are scads of good books / web pages for both. Also, your
> colleagues and professors. People are better than books.
>
> If you're thinking of being a normally-numerate to somewhat-high-tech
> political scientist, you can expect that:
>
> (1) You might well be able to do a fair chunk of your work in SPSS.
> (2) You will need routines (estimators, corrections) that SPSS
> doesn't have, and probably sooner rather than later.
> (3) At that point, you need to turn to the right tool for the job.
> Maybe it'll be R or Stata, or a routine written for one or the
> other. Or maybe it will be any of various pieces of boutique
> software.
> (4) You're going to end up learning at least the basics of R and
> Stata. And several pieces of boutique software relevant to your
> dissertation research. This is not a matter of choice except
> insofar as you choose to be in graduate school. Learning these
> tools will be imposed on you by people who insist that you
> use the
> right tool for the job, when those right tools are scattered
> across different programs and operating systems. This is normal.
> Resistance to this idea is also normal, but, well, resistance is
> futile.
>
> I guess what I mean is that you don't need to worry about which is
> bestand which you should learn. You win the jackpot -- you get to
> learnthem all.
>
> Best,
>
> Jim Battista
> [log in to unmask]
>
> **********************************************************
> Political Methodology E-Mail List
> Editor: Karen Long Jusko <[log in to unmask]>
> **********************************************************
> Send messages to [log in to unmask]
> To join the list, cancel your subscription, or modify
> your subscription settings visit:
>
> http://polmeth.wustl.edu/polmeth.php
>
> **********************************************************
>
**********************************************************
Political Methodology E-Mail List
Editor: Karen Long Jusko <[log in to unmask]>
**********************************************************
Send messages to [log in to unmask]
To join the list, cancel your subscription, or modify
your subscription settings visit:
http://polmeth.wustl.edu/polmeth.php
**********************************************************
|