Dear Yves,
I've switched to Stata from Spss about ten years ago. I still use
Spss, sometimes, as we have a campus license. However my dept. as a
whole is currently re-orienting its preferences toward Stata.
As regards costs, we are not just concerned about our own costs. We'd
also like to train our students with a package that will be affordable
for whatever organization will employ them after graduation. Stata is
definitely a better value for money.
My experience is limited: data & files management, descriptive
statistics, regression models for linear and categorical variables,
survival analysis, producing outputs for papers. So I probably use
Stata at 5% of its capabilities. You can have more suggestions from
more experienced users
Some elements - according to my experience- you may wish to consider:
- Spss (especially after rel 12) is rather baroque and use more
hardware resources than Stata does
- I find rather annoying, and a source of errors, in Spss (after rel
12) the possibility to work on more datasets at the same time; I'd
prefer Stata approach, allowing to open simultaneously different
sessions of the program for different jobs.
- programming in Stata is less time consuming than in Spss
- the (supposed to be) better user interface in Spss is
counterbalanced by the possibility that unaware users may get some
results without any understanding of what they get. The improved user
interface in Stata still require the user to think about his/her
choices.
- I've not tested this on statistics computations, but it happened to
me to measure performances on data transformations (strings -> unique
number IDs for a large dataset) and Stata: Spss time needed was about
1:10
- generating output for reports etc. is less straightforward in Stata,
but several ado files have been created (outreg, tabout, ...) to deal
with this and to better customize your outputs.
- last but not least, I probably consider the Stata community (sharing
add-ons and knowledge) the main reason for being "Statish"
At the end, what I suggest you is to organize a "match" Spss-Stata, by
considering a typical data analysis work in your organization. Develop
it with the two packages and discuss the pros and the cons of the 2
options.
Finally, you'll have in any case to consider that people need time and
support in order to switch from one package to another.
Hope this helps
Teresio
____________________________________________________
dr. Teresio Poggio
LaboR - Dipartimento di Sociologia e ricerca sociale
Università degli studi di Trento
Via Verdi, 26
38100 Trento, Italy
Tel +39 0461/881406
fax: +39 0461/881348
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
