A LOT depends on what kind of analyses you wish to do with the software, the
capability level of the user w.r.t. statistics, AND the frequency with which
you'll use the software (related to "relearning" how to use the software
after an extended absence).
I've used a bunch of software in my day. I used a LOT of SAS when I worked
for General Motors. Where I work now, I used SAS for a few years mores, we
used StatGraphics (both DOS & Windows) for four or five years; we used JMP
for three years, and now have been using MINITAB for three years I have no
experience with S-plus or SPSS, but did do a lot of benchmarking 3-4 years
ago when we were looking for an alternative to JMP (when they stopped
offering concurrent licensing). At that point I spent a lot of time looking
at Statistica, but didn't spend much time with SPSS since it had a weak DOE
tool. At the time SysStat's DOE tool was also pretty archaic, so I didn't
spend much time on that.
Here are my slightly biased thoughts:
SAS - VERY POWERFUL, but hard to learn except for the basics. If you don't
need to tweak too many of the options the Procs can kick out lots of stuff
pretty easily. Working with data sets and graphs, I felt, was a major pain.
I spent a lot of time on Excel to avoid SAS. Its also very expensive:
$1000+/year.
S+ - I hear it is becoming very popular with the stat development bunch. It
is easier to do new neat stuff with S+ than with SAS. It appears to be more
complicated than some others, but as I said I don't have much exposure to
S+.
SPSS - not a lot of experience here. It appears to be focused more on
regression analysis or something than the industrial stats world in which I
live.
MINITAB - easy to use, clean menus, relatively cheap, surprisingly powerful.
Nice downloadable fully functional demo. Best for the occasional user. My
favorite. Oh, and it has the absolute best help system I've ever seen in any
software. They did real good with that. Oh, and all manuals are provided in
a nicely done Acrobat format. It has some quirks when trying to edit
graphs - you have to remember that they are created via macros, so if you
want to modify something on the graph you basically have to create a whole
new graph using additional options rather than just modify the existing
graph. That will all be changed with version 14 coming out in a year or so.
Did I mention it's reliability and accelerated testing tool is just about
the best I've seen?
JMP - powerful, claims to be easy to use, but I always found it quirky. For
example, ALL of the modeling is done via its "Fit Model" tool. To do
anything you'd better know this well I think MINITAB is a better choice. JMP
3's help and documentation was pretty mediocre. Also it is really a
Macintosh program sitting in Windows. It has kept the Mac quirkiness.
StatGraphics - also very powerful with some nice tools, BUT was cursed with
a rather odd looking (just visually) windows environment. They now have a
new version out - I'd be curious to see if they improved it. If they did it
would be a real contender with MINITAB.
Statistica - I found this to be very powerful, but I had to spend a lot of
time learning it. It really wasn't very intuitive in many areas. It has a
very nice graphics engine. It was also pretty expensive. How can they say
they've won every comparison?
As mentioned by someone else, MatLab and Mathematica do have some statistics
content, but they are no more stat packages than is Excel. Which brings up
the semi-famous Excel math errors - be very careful with Excel.
My advice, get the demos and spend lots of time playing with them. MINITAB's
demo is fully functional - it just times out after 30 days (you can call
them to extend it to 60 days). JMP's demo is pretty crippled - you have to
use their data without any hint which methods use which data sets - it
drives me nuts (most of my company uses JMP, but I can't stand it anymore
now that I have MINITAB). Statistica used to give you demo with just a few
capabilities - it was pretty hard to get the feel with. I think I read that
StatGraphics now has a fully functional demo like MINITAB.
If you find yourself scratching your head over the demo, then maybe its too
hard to use - it shouldn't have to be. After I'd spent weeks playing with
Statistica I finally loaded the MINITAB demo - the software's menus were so
clean and logical it was literally like a breath of fresh air. I'm not
kidding.
I'd look at them in this order: MINITAB, StatGraphics, JMP, Statistica, and
then the others. I think SPSS would be next and then S-plus, but don't
really know about those. I've heard some say MINITAB is just for teaching -
don't be fooled - it is just as powerful as the others, maybe more so.
"Bob Hayden" <[EMAIL PROTECTED]> wrote in message
[EMAIL PROTECTED]">news:[EMAIL PROTECTED]...
> ----- Forwarded message from Jan -----
>
> Could you please give advantages the major stat. systems
> over others and strength and weakness of them.
> Eg. could you compare SAS, S (S+), SPSS, Minitab, Statgraphics (plus),
> Matlab, Maple, Mathematica?
>
> ----- End of forwarded message from Jan -----
>
> Matlab, Maple, are Mathematica mathematics programs with some
> stats. cabability. They are not very easy to learn. I would not use
> them with beginners who are not mathematics majors.
>
> Minitab is easy to use and there is a huge amount of documentation.
> Many statistics textbooks in the US include info on Minitab.
>
> I think SAS and S are not neraly as easy to use.
>
> _
> | | Robert W. Hayden
> | | Work: Department of Mathematics
> / | Plymouth State College MSC#29
> | | Plymouth, New Hampshire 03264 USA
> | * | fax (603) 535-2943
> / | Home: 82 River Street (use this in the summer)
> | ) Ashland, NH 03217
> L_____/ (603) 968-9914 (use this year-round)
> Map of New [EMAIL PROTECTED] (works year-round)
> Hampshire http://mathpc04.plymouth.edu (works year-round)
>
> The State of New Hampshire takes no responsibility for what this map
> looks like if you are not using a fixed-width font such as Courier.
>
> "Opportunity is missed by most people because it is dressed in
> overalls and looks like work." --Thomas Edison
>
>
>
> =================================================================
> Instructions for joining and leaving this list and remarks about
> the problem of INAPPROPRIATE MESSAGES are available at
> http://jse.stat.ncsu.edu/
> =================================================================
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
