RE: st: Spss vs Stata

["Ploutz-Snyder, Robert (JSC-SK)[USRA]" <robert.ploutz-snyder-1@nasa.gov>]

I will chime in also, as an SPSS user of 20 years...now converted almost entirely to Stata (except for ANOVA).

While I will agree that SPSS has pretty output, I am also reminded of the old output (thinking SPSS version 4 ish), and the "wow" factor that we SPSS users experienced when they went to the "new" format that was much prettier.  Yes, it was nice looking, but those of us who made that transition from SPSS output that looked at lot like today's Stata to the now pretty version also remember the cost.  It wasn't severe then, but run practically any routine in SPSS today and it takes much longer to produce that pretty output.   Run something modestly complicated and you might as well go get a cup of coffee.  Try something akin to a 2 layer xtmixed and you may have to let your SPSS sit overnight.  I'm not joking, and I have 64bit whiz-bang machine.   Aside from processing speed, SPSS is a complete memory hog that will crash all but the top-end machines if you are a multi-tasker.  And then, of course, there the $$$ cost.

Yes, SPSS output looks nice.  But even with that nice output it's not publication quality so you STILL end up getting that output to your word processor of choice (or LaTEX) to format for manuscript submission.  Maybe some are willing to dedicate the time to SPSS output modification, but personally I found that more frustrating than a simple copy/paste. 

So for my 2-cents (and the Stata monitors), I would hope that Stata would NOT dedicate much effort in making pretty output.  What was and continues to be refreshing to me as a relatively new Stata user is the SPEED that you can get things done.  The syntax language is intuitive and brief (compared to SPSS and SAS), and it runs routines very fast. The output isn't pretty but it gives you what you need.   True, it would be nice to have  better copy/paste functionality going from Stata to Excel or Word, but aside from that, I don't really need all that pretty output from my statistics package.  Let's face it--we don't rely on Excel to do our statistics, so why rely on Stat to do publication quality tables?  

As for graphics, Stata's graphics commands are more difficult to master but the end result is far better than SPSS graphics.  Mike Mitchell's "Visual Guide to Stata Graphics" is a must-have for newbie Stata users, as it gets you up and running quickly for making very nice graphics.  Before long, you can tweak your code to publication quality...

What SPSS still maintains over Stata is better ANOVA routines, particularly Repeated-Measures fixed-factor designs.  Stata treats RM designs a bit strangely, I believe because it seems to "wrap" ANOVA code around Regression methods.  It's non-intuitive and can provide results that aren't typical of RM ANOVA (consider how it uses full-n for fixed-factor RM ANOVA without listwise elimination of subjects who are missing an observation).  I would much prefer to see Stata invest in re-working their ANOVA code and analyses so that it is more consistant with SAS or SPSS methodologies, offers more in terms of assumption testing (ex. Sphericity tests), and is more intuitive.

But pretty output--that's not high on my priority list.  

Rob



-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of E. Paul Wileyto
Sent: Sunday, August 01, 2010 11:37 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Spss vs Stata

My two cents.  Graphics is the only bad aspect of Stata, but SPSS is 
almost as rudimentary in graphics.  If you want publication quality 
graphics quickly, the only package to consider is SigmaPlot.  I do 
analysis in Stata, quickly collapse and reshape data for graphs, and 
copy the reshaped data into SigmaPlot.  Excel cannot match any of 
SigmaPlot's capabilities, and it exports to EPS, PDF, or JPG.  Also, you 
can copy & paste figures without background directly to powerpoint for a 
nice colored background.

Paul

On 8/2/2010 12:14 AM, Stas Kolenikov wrote:
> SPSS, especially these days as it has become the "predictive
> analytics" software rather than a statistical package for social
> sciences, seems to operate from the viewpoint of "I click here, I
> click there, and I drag this into a report -- voila". Stata has
> traditionally been oriented at more meditative researchers in academic
> settings who are less concerned with flashy presentations, as that's
> rarely a selling point of a research paper. Also, SPSS is only good
> under Windows, while Stata takes a lot of (well substantiated) pride
> in working under all common operating systems.
>
> A number of tools exist to go from Stata output to other external
> programs, such MS Word/Excel, on one hand, and LaTeX, on the other. I
> am certainly in the latter camp, and I get everything I need from
> -estout- and low level -file- commands with which I write my results.
> Of course there are times when I have to do some copy/paste, and
> that's less convenient than from SPSS. More MS-specific tools are
> available with -outreg-, -xmltab-, and probably five or so others.
>
> On Sun, Aug 1, 2010 at 7:04 PM, Yves Therriault<ytherriault@gmail.com>  wrote:
>    
>> Dear Stata users,
>>
>> I've been using SPSS for 20 years. For the time being, I still use a
>> six years old version of SPSS (12.01). For many reasons, I told my
>> organisation that I would rather purchasing one licence of Stata
>> instead of upgrading to the newest release of SPSS (18.0).
>>
>> A few months ago, I've asked former SPSS users to write about their
>> experience with Stata. Generally, people who have decided to switch
>> from SPSS to Stata aren't looking back.
>>
>> In order to learn the basics of Stata more easily when the software is
>> installed on my computer, I bought 3 introductory books : A Gentle
>> Introduction to Stata ; An Introduction to Stata for Health
>> Researchers and, finally Data Analysis Using Stata.
>>
>> I'm interested to hear from those who made the switch to Stata and
>> particularly about the lurning curve regarding the management of Stata
>> output. Generally, Stata seems to be far more superior to SPSS in many
>> ways. But, in my humble opinion, Stata is currently far behind SPSS
>> concerning how it displays its output. I'm aware that a lot of
>> routines have been written by Stata users to customize outputs though.
>>
>> Just in case Stata programmers were monitoring this list, I would be
>> very pleased if the company decided to "improve" the output display in
>> a future release. Perhaps, another possible feature would gave the
>> user the choice between a "standard" Stata output and a Spss-like
>> output.
>>      
>    

-- 
E. Paul Wileyto, Ph.D.
Assistant Professor of Biostatistics
Tobacco Use Research Center
School of Medicine, U. of Pennsylvania
3535 Market Street, Suite 4100
Philadelphia, PA  19104-3309

215-746-7147
Fax: 215-746-7140
epw@upenn.edu

*
*   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/

*
*   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/