STATA

[Stan Gorodenski]

I have looked into STATA as an alternative to SPSS. The statistics it
has appears to be every bit as good as SPSS's. Even its ARIMA procedure
is more flexible and more powerful like the original PACK program than
SPSS. For years SPSS has not brought up its ARIMA procedure to where it
should be, and this still hasn't happened under IBM (I don't think). Its
programming capabilities, although it would require learning a lot of
new syntax, appear to be just as good and maybe even better than SPSS's.
Does anyone have any experience with STATA to verify this and to say
where STATA might fall short compared to SPSS, and be a disappointment?
Stan

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[SR Millis-3]

Please note:

Stata is an invented word, not an acronym, and should not appear with all letters capitalized: please write “Stata”, not “STATA”.

http://www.stata.com/support/faqs/resources/statalist-faq/#spell

I love Stata.  Once you buy it, you own it.  It does more than SPSS.  It comes with its own SEM program and missing data programs.  Incredible customer support.  Active user community.

 

~~~~~~~~~~~
Scott R Millis, PhD, ABPP, CStat, PStat®
Board Certified in Clinical Neuropsychology, Clinical Psychology, & Rehabilitation Psychology 
Professor
Wayne State University School of Medicine
Email: [hidden email]
Email: [hidden email]
Tel: [phone]

---

From: Stan Gorodenski <[hidden email]>
To: [hidden email]
Sent: Friday, May 3, 2013 2:23 AM
Subject: STATA

I have looked into STATA as an alternative to SPSS. The statistics it
has appears to be every bit as good as SPSS's. Even its ARIMA procedure
is more flexible and more powerful like the original PACK program than
SPSS. For years SPSS has not brought up its ARIMA procedure to where it
should be, and this still hasn't happened under IBM (I don't think). Its
programming capabilities, although it would require learning a lot of
new syntax, appear to be just as good and maybe even better than SPSS's.
Does anyone have any experience with STATA to verify this and to say
where STATA might fall short compared to SPSS, and be a disappointment?
Stan

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Marta Garcia-Granero]

El 03/05/2013 8:23, Stan Gorodenski escribió:
> I have looked into STATA as an alternative to SPSS. The statistics it
> has appears to be every bit as good as SPSS's. Even its ARIMA procedure
> is more flexible and more powerful like the original PACK program than
> SPSS. For years SPSS has not brought up its ARIMA procedure to where it
> should be, and this still hasn't happened under IBM (I don't think). Its
> programming capabilities, although it would require learning a lot of
> new syntax, appear to be just as good and maybe even better than SPSS's.
> Does anyone have any experience with STATA to verify this and to say
> where STATA might fall short compared to SPSS, and be a disappointment?

Hi Stan:

As a teacher, I had to convert every single SPSS procedure I use in my
classes to Stata. My students (graduated: researchers, doctorate,
master...) complain a lot about results hanling (they miss the pivot
tales, and so do I).

I've had a hard time with ANOVA in Stata. To be more precise, with
multivariate tests in repeated measures ANOVA as an alternative to
epsilon corrected significance. The process with Stata was a bit (mmm,
rather, mmm... A LOT) more complicated with Stata than with SPSS.
Besides, trying to convince short sighted (metaphorically speaking)
researchers that Mauchly's sphericity test is really not that good and
is not necessary (Stata doesn't calculate it) was difficult. SPSS is
more... user friendly when ANOVAS are involved, although Stata's xtmixed
is really good.

Also, classic non parametric tests (not that they are very used today,
are they?) are poorer, with even a bug in Wilcoxon's signed rank test
when dealing with non integer data that needs a tedious workaround.

For the rest, Stata is fantastic, and, in some aspects (ROC or survival
analysis, for instance) clearly superior to SPSS.

Sorry, ibm dot com (formerly spss dot com) people. I had to be honest.

But I still miss SPSS (I knew by heart many commands and the GUI, I
don't work as fast with Stata).

Best wishes,
Marta GG

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

I
don't work as fast with Stata).

And this from a very experienced cracker jack statistician!

This is what I hear from many people who have been forced to change to STATA for short term budgetary reasons.

In the early days a certain large computer company that focused on batch business applications had signs something like
"Machines are expensive. People are cheap. Bench check your work."

Today my experience is that "Machines are cheap. People are expensive.  Human factors are critical. Total cost of ownership must be considered." YMMV.
On 5/3/2013 7:16 AM, Marta García-Granero-2 [via SPSSX Discussion] wrote:

I
don't work as fast with Stata).

[Bruce Weaver]

"And this from a very experienced cracker jack statistician!"

Let's be fair here, Art.  When someone is an advanced user of a stats package (like Marta with SPSS), they use a huge store of implicit knowledge.  They only truly realize how much implicit knowledge they've been using when they switch to a new, unfamiliar package.  So a big part of the reason Marta works more slowly with Stata is that she is so well-versed in SPSS.  But if we turned the tables, and asked a Stata wizard to switch to SPSS (or any other package), they would also work more slowly.  

This is not to say that all packages are equally good for all tasks.  For example, Marta confirmed what that fellow from the UCLA Statistical Consulting service (was it Mitchell?) wrote a few years ago in a piece comparing the major stats packages -- i.e., he said that Stata was fantastic for regression models, but not as good as SPSS for traditional ANOVA models, especially those involving repeated measures.

One of the problems with these stats packages comparisons is that most of the opinions expressed come from people who use one of the packages almost exclusively.  Out come the straw men, therefore, and it devolves into arguments of the form, "My Dad can whip your Dad!", etc.  That's what I really liked about Mitchell's piece -- he clearly had a lot of experience working with all of the packages he was commenting on.

Marta, what can you tell us about basic data management with Stata?  

Cheers,
Bruce

Art Kendall wrote

I
       
        don't work as fast with Stata).
     
      And this from a very experienced cracker jack
          statistician!
       
      This is what I hear from many people who have been forced to
      change to STATA for short term budgetary reasons.
     
      In the early days a certain large computer company that focused on
      batch business applications had signs something like
      "Machines are expensive. People are cheap. Bench check your work."
     
      Today my experience is that "Machines are cheap. People are
      expensive.  Human factors are critical. Total cost of ownership
      must be considered." YMMV.
      Art Kendall
Social Research Consultants
      On 5/3/2013 7:16 AM, Marta García-Granero-2 [via SPSSX Discussion]
      wrote:
   
     I
     
      don't work as fast with Stata).

[Dale]

For several years I have been using SYSTAT as my base of operations with SPSS as one of the supplemental.  For what I need most of the time it is just a better alternative.   There are things SPSS currently does very well relative to SYSTAT, but the need to add-on modules, rental rather than own type licensing, and other issues have just not been a positive for me.  It may be something to look into as well.

Dale


Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497



-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Stan Gorodenski
Sent: Friday, May 03, 2013 1:23 AM
To: [hidden email]
Subject: STATA

I have looked into STATA as an alternative to SPSS. The statistics it has appears to be every bit as good as SPSS's. Even its ARIMA procedure is more flexible and more powerful like the original PACK program than SPSS. For years SPSS has not brought up its ARIMA procedure to where it should be, and this still hasn't happened under IBM (I don't think). Its programming capabilities, although it would require learning a lot of new syntax, appear to be just as good and maybe even better than SPSS's.
Does anyone have any experience with STATA to verify this and to say where STATA might fall short compared to SPSS, and be a disappointment?
Stan

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

Most users need to come up to speed on both stat and software.
Marta also moved to STATA a few years ago.  Yes any new package has a learning curve. 
the first package takes a lot longer than the second. The second somewhat longer than the third and so forth.  Each "old" package facilitates learning the next. Much of the implicit knowledge is at least similar. Implicit knowledge: knowing the stat behind the procedure, what questions it addressees, knowing what a file is,  etc., etc., facilitates learning the new package.    I do not believe that knowledge of one package impedes learning others. 
Analogously, I found Latin very helpful in learning French and Greek.

YMMV but I find that teaching/mentoring  stat involves more than just some stat concepts.  There are things about data handling, reporting, computer concepts, etc. that are also new. I can remember people needing clarification about what is a row and what is a column.

I would also like to hear from Marta about the human factors: consistency and readability of syntax etc, especially on the part of an analysis that takes up most of the clock time.  I would also like to hear how she perceives her student's experience. Of course people who follow list this know that I consider the human factors to be of major importance.

People I have talked to who learned other packages first and SPSS later.  They grumbled at first about the new ways of doing things, but after a while said they were glad to have added SPSS to their repertoire.  Maybe they were shining me on but they said that for all the data prep, documentation, quality assurance review, etc.  SPSS was more efficient for them.

Also you know that I have often said that for an organization it should not be { SPSS xor STATA xor IMSL xor SAS }.
xor is a logical operator from FORTRAN meaning A or B but not A and B

BTW SPSS was not my first package. Before SPSS I used SIFT, APL pack, BMDP and many standalone or ad hoc programs.
On 5/3/2013 9:06 AM, Bruce Weaver [via SPSSX Discussion] wrote:

"And this from a very experienced cracker jack statistician!"

Let's be fair here, Art.  When someone is an advanced user of a stats package (like Marta with SPSS), they use a huge store of implicit knowledge.  They only truly realize how much implicit knowledge they've been using when they switch to a new, unfamiliar package.  So a big part of the reason Marta works more slowly with Stata is that she is so well-versed in SPSS.  But if we turned the tables, and asked a Stata wizard to switch to SPSS (or any other package), they would also work more slowly.  

This is not to say that all packages are equally good for all tasks.  For example, Marta confirmed what that fellow from the UCLA Statistical Consulting service (was it Mitchell?) wrote a few years ago in a piece comparing the major stats packages -- i.e., he said that Stata was fantastic for regression models, but not as good as SPSS for traditional ANOVA models, especially those involving repeated measures.

One of the problems with these stats packages comparisons is that most of the opinions expressed come from people who use one of the packages almost exclusively.  Out come the straw men, therefore, and it devolves into arguments of the form, "My Dad can whip your Dad!", etc.  That's what I really liked about Mitchell's piece -- he clearly had a lot of experience working with all of the packages he was commenting on.

Marta, what can you tell us about basic data management with Stata?  

Cheers,
Bruce

Art Kendall wrote

I
       
        don't work as fast with Stata).
     
      And this from a very experienced cracker jack
          statistician!
       
      This is what I hear from many people who have been forced to
      change to STATA for short term budgetary reasons.
     
      In the early days a certain large computer company that focused on
      batch business applications had signs something like
      "Machines are expensive. People are cheap. Bench check your work."
     
      Today my experience is that "Machines are cheap. People are
      expensive.  Human factors are critical. Total cost of ownership
      must be considered." YMMV.
      Art Kendall
Social Research Consultants
      On 5/3/2013 7:16 AM, Marta García-Granero-2 [via SPSSX Discussion]
      wrote:
   
     I
     
      don't work as fast with Stata).

---

If you reply to this email, your message will be added to the discussion below:

http://spssx-discussion.1045642.n5.nabble.com/STATA-tp5719891p5719907.html

To start a new topic under SPSSX Discussion, email [hidden email]
To unsubscribe from SPSSX Discussion, click here.
NAML

[Allan T. Walters]

I have used both Stata, SAS, and SPSS. I have used SPSS rather extensively since I was a graduate student, some forty-five years, or so, ago.  I have only used Stata sporadically in the last twenty or so years.  I mostly just write the code; in Stata I have never used the graphical interface.  I have also used Stata primarily to analyze SCF data/and or done imputations with other data sets.  I have translated a fifty page SAS code file to Stata with not a great deal of difficulty.  I find that having used SPSS for so many years that it is easier and quicker to do simple things, like simple data management, in SPSS and have used the graphical interface mode which is pretty well implemented.  But, with a little relearning, I have use the survey module in Stata rather extensively with not much difficulty. Stata has rather extensive facility for producing graphical output. There are actually paper Stata manuals which are relatively easy to use and fairly extensive in their coverage, though somewhat costly.  Documentation for SPSS is now, I believe, solely in PDF form. For those in a University setting who wish to purchase a perpetual license, I believe the price of SPSS is much more expensive the price of Stata.  Students, and I presume faculty, get a pretty good break in costs for both SPSS and SAS with a must-have yearly renewal cycle. As opposed to SPSS, all modules in Stata come with base Stata, no adding additional modules as one must in SPSS.  SPSS has a yearly upgrade cycle; Stata has an every other year renewal cycle. I would prefer to work in Stata, though I can understand that it has a bit of a learning curve, and does some things very differently when compared to SPSS and SAS.  Actually some aspects of coding in SAS are more elegant and concise than can be done in Stata. But, alas SAS does not work on the Mac and I do not believe you can purchase a perpetual license even in the Microsoft operating system world. 

Yes machines are relatively inexpensive, but yearly renewal costs add up, in at least my experience.  

This at least represents my experience, others may have more recent experience with the pricing on SPSS.  By the way, some years ago, I also ran SPSS (with some difficulty) under both IBM Unix and DEC Ultrix, and I believe a perpetual license was then quite reasonable.

Allan T. Walters

[hidden email]

On May 3, 2013, at 9:54 AM, Art Kendall <[hidden email]> wrote:

Most users need to come up to speed on both stat and software.
Marta also moved to STATA a few years ago.  Yes any new package has a learning curve. 
the first package takes a lot longer than the second. The second somewhat longer than the third and so forth.  Each "old" package facilitates learning the next. Much of the implicit knowledge is at least similar. Implicit knowledge: knowing the stat behind the procedure, what questions it addressees, knowing what a file is,  etc., etc., facilitates learning the new package.    I do not believe that knowledge of one package impedes learning others. 
Analogously, I found Latin very helpful in learning French and Greek.

YMMV but I find that teaching/mentoring  stat involves more than just some stat concepts.  There are things about data handling, reporting, computer concepts, etc. that are also new. I can remember people needing clarification about what is a row and what is a column.

I would also like to hear from Marta about the human factors: consistency and readability of syntax etc, especially on the part of an analysis that takes up most of the clock time.  I would also like to hear how she perceives her student's experience. Of course people who follow list this know that I consider the human factors to be of major importance.

People I have talked to who learned other packages first and SPSS later.  They grumbled at first about the new ways of doing things, but after a while said they were glad to have added SPSS to their repertoire.  Maybe they were shining me on but they said that for all the data prep, documentation, quality assurance review, etc.  SPSS was more efficient for them.

Also you know that I have often said that for an organization it should not be { SPSS xor STATA xor IMSL xor SAS }.
xor is a logical operator from FORTRAN meaning A or B but not A and B

BTW SPSS was not my first package. Before SPSS I used SIFT, APL pack, BMDP and many standalone or ad hoc programs.
On 5/3/2013 9:06 AM, Bruce Weaver [via SPSSX Discussion] wrote:

"And this from a very experienced cracker jack statistician!"

Let's be fair here, Art.  When someone is an advanced user of a stats package (like Marta with SPSS), they use a huge store of implicit knowledge.  They only truly realize how much implicit knowledge they've been using when they switch to a new, unfamiliar package.  So a big part of the reason Marta works more slowly with Stata is that she is so well-versed in SPSS.  But if we turned the tables, and asked a Stata wizard to switch to SPSS (or any other package), they would also work more slowly.  

This is not to say that all packages are equally good for all tasks.  For example, Marta confirmed what that fellow from the UCLA Statistical Consulting service (was it Mitchell?) wrote a few years ago in a piece comparing the major stats packages -- i.e., he said that Stata was fantastic for regression models, but not as good as SPSS for traditional ANOVA models, especially those involving repeated measures.

One of the problems with these stats packages comparisons is that most of the opinions expressed come from people who use one of the packages almost exclusively.  Out come the straw men, therefore, and it devolves into arguments of the form, "My Dad can whip your Dad!", etc.  That's what I really liked about Mitchell's piece -- he clearly had a lot of experience working with all of the packages he was commenting on.

Marta, what can you tell us about basic data management with Stata?  

Cheers,
Bruce

Art Kendall wrote

I
       
        don't work as fast with Stata).
     
      And this from a very experienced cracker jack
          statistician!
       
      This is what I hear from many people who have been forced to
      change to STATA for short term budgetary reasons.
     
      In the early days a certain large computer company that focused on
      batch business applications had signs something like
      "Machines are expensive. People are cheap. Bench check your work."
     
      Today my experience is that "Machines are cheap. People are
      expensive.  Human factors are critical. Total cost of ownership
      must be considered." YMMV.
      Art Kendall
Social Research Consultants
      On 5/3/2013 7:16 AM, Marta García-Granero-2 [via SPSSX Discussion]
      wrote:
   
     I
     
      don't work as fast with Stata).

---

If you reply to this email, your message will be added to the discussion below:

http://spssx-discussion.1045642.n5.nabble.com/STATA-tp5719891p5719907.html

To start a new topic under SPSSX Discussion, email <a href="x-msg://1062/user/SendEmail.jtp?type=node&amp;node=5719914&amp;i=1" target="_top" rel="nofollow" link="external">[hidden email]
To unsubscribe from SPSSX Discussion, <a moz-do-not-send="true" href="x-msg://1062/" target="_top" rel="nofollow" link="external">click here.
NAML

---

View this message in context: Re: STATA
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

[John F Hall]

. . . and has anyone calculated the institutional cost in staff time (and stress) for the changeover? 

 

Has anyone actually displayed SPSS and Stata syntax side-by-side for the same basic jobs, data input, data dictionary, transformations, tabulation and graphics?

 

I’ve been retired these 20 years or so, but I’d still stick to SPSS (and try to haggle a bigger discount for academic sites).

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

Start page:  www.surveyresearch.weebly.com/spss-without-tears.html

  

 

 

 

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
Sent: 03 May 2013 13:39
To: [hidden email]
Subject: Re: STATA

 

I
don't work as fast with Stata).

And this from a very experienced cracker jack statistician!

This is what I hear from many people who have been forced to change to STATA for short term budgetary reasons.

In the early days a certain large computer company that focused on batch business applications had signs something like
"Machines are expensive. People are cheap. Bench check your work."

Today my experience is that "Machines are cheap. People are expensive.  Human factors are critical. Total cost of ownership must be considered." YMMV.

Art Kendall

Social Research Consultants

On 5/3/2013 7:16 AM, Marta García-Granero-2 [via SPSSX Discussion] wrote:

I
don't work as fast with Stata).

 

Art Kendall
Social Research Consultants

 

---

View this message in context: Re: STATA
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

[Stan Gorodenski]

I am away from my pc that has all the other responses to my question about
Stata. I was going to wait to get back to respond to one of them, but I
will do so now. Someone didn't know how the cost of SPSS compared to
Stata. Way back when ver 18 of SPSS was coming out (I think that was
around the time) it would have cost me about $2,000 to upgrade to ver 18.
I waffled for some time and during that time IBM bought out SPSS. Now to
upgrade to the current version (which I know is about 3 versions newer
than 18) it would cost me about $10,000 to get  all the capabilities I
currently have. Stata would cost $2,500 and it includes *all* modules,
compared to SPSS where one has to buy them individually. What would cost
me $10,000 from SPSS does not include ARIMA, but the Stata package does
and its procedure appears to be more powerful than SPSS's. Recently I had
asked an IBM employee if SPSS had improved the ARIMA procedure from what
it had been, and removed some of its limitations, but the employee could
not answer my question.

Some time ago I had asked the group whether ver 15 would run in Win 7 and
on a 64-bit machine. The responses I got said it would. I have recently
purchased a Lenovo laptop that has Win 7 Professional 64-bit as the
operating system. I am currently working on it to make sure all the other
programs I need will run on it. Then I will install SPSS. If I cannot get
another license code (which I would have been able to without a problem
before IBM took over) then I would have to decide on whether to buy SPSS
for its exhorbitant IBM cost or get Stata. I am not afraid of learning a
new programming language. In some ways I like to work with strange
symbols, but this does not mean I am proficient in programming. What would
concern me is the environment, or whatever it might be called, Stata runs
in that might make it annoying and more cumbersome to use (for example,
how does it save and read system files; getting, viewing, modifying,
copying parts of and saving output?).

I looked into Stata out of curiosity because someone on this group
mentioned Stata is an alternative and became surprised at what seemed to
be an equal package, maybe even better, compared to SPSS. Currently, it
appears the Statagroup (If I recall the name correctly) is offering and
supporting their product like SPSS used to before IBM took over. I wonder
how long that will last. Maybe SPSS couldn't survive and had to sell out
to IBM because of its business model whch to me includes the manner of
licensing and cost.
Stan

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Mark Miller]

Comparing SPSS and Stata is like comparing oranges to apples (or vegetables).

Stata (like R and Matlab) is (primarily) an in-memory processing package, data size is limited to RAM.

SPSS (like SAS) is fundamentally designed to handle data far larger than RAM.

Both packages have advantages and disadvantages (which is which depends on your point of view).

Compiling a comparative list is a useful exercise, but such work may have a

short lifetime as all of the packages change every year (or sooner)

and it is common for different compilers to generate quite different lists.

Compare for example these two comparisons

    http://stanfordphd.com/Statistical_Software.html

and

   http://brenocon.com/blog/2009/02/comparison-of-data-analysis-packages-r-matlab-scipy-excel-sas-spss-stata/

Robert  A. Muenchen (University of Tennessee) produced a very thorough comparison

    https://science.nature.nps.gov/im/datamgmt/statistics/R/documents/R_for_SAS_SPSS_users.pdf

which eventually blossomed into two books published by Springer 

   

R for SAS and SPSS Users (Statistics and Computing) by Robert A. Muenchen (Jul 26, 2011

R for Stata Users (Statistics and Computing) by Robert A. Muenchen and Joseph M. Hilbe (Jul 1, 2012

... Mark Miller

On Sat, May 4, 2013 at 9:55 AM, Stanley Gorodenski <[hidden email]> wrote:

I am away from my pc that has all the other responses to my question about
Stata. I was going to wait to get back to respond to one of them, but I
will do so now. Someone didn't know how the cost of SPSS compared to
Stata. Way back when ver 18 of SPSS was coming out (I think that was
around the time) it would have cost me about $2,000 to upgrade to ver 18.
I waffled for some time and during that time IBM bought out SPSS. Now to
upgrade to the current version (which I know is about 3 versions newer
than 18) it would cost me about $10,000 to get  all the capabilities I
currently have. Stata would cost $2,500 and it includes *all* modules,
compared to SPSS where one has to buy them individually. What would cost
me $10,000 from SPSS does not include ARIMA, but the Stata package does
and its procedure appears to be more powerful than SPSS's. Recently I had
asked an IBM employee if SPSS had improved the ARIMA procedure from what
it had been, and removed some of its limitations, but the employee could
not answer my question.

Some time ago I had asked the group whether ver 15 would run in Win 7 and
on a 64-bit machine. The responses I got said it would. I have recently
purchased a Lenovo laptop that has Win 7 Professional 64-bit as the
operating system. I am currently working on it to make sure all the other
programs I need will run on it. Then I will install SPSS. If I cannot get
another license code (which I would have been able to without a problem
before IBM took over) then I would have to decide on whether to buy SPSS
for its exhorbitant IBM cost or get Stata. I am not afraid of learning a
new programming language. In some ways I like to work with strange
symbols, but this does not mean I am proficient in programming. What would
concern me is the environment, or whatever it might be called, Stata runs
in that might make it annoying and more cumbersome to use (for example,
how does it save and read system files; getting, viewing, modifying,
copying parts of and saving output?).

I looked into Stata out of curiosity because someone on this group
mentioned Stata is an alternative and became surprised at what seemed to
be an equal package, maybe even better, compared to SPSS. Currently, it
appears the Statagroup (If I recall the name correctly) is offering and
supporting their product like SPSS used to before IBM took over. I wonder
how long that will last. Maybe SPSS couldn't survive and had to sell out
to IBM because of its business model whch to me includes the manner of
licensing and cost.
Stan

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Enzmann]

Here it is:
http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0310&L=spssx-l&P=R14717

BTW: "STATA" is no acronym and should be spelled "Stata".

I used SPSS extensively for about 20 years. For a while I experimented
with R to solve problems that could not be solved using SPSS, but I gave
up - not completely, though - because R is far less user friendly than
SPSS or Stata. Finally I made a move to Stata. The last 5 years I am
using it almost exclusively and I am happy that I decided to make the
move: Not only it is less expensive, at the same time it offers far more
and better capabilities of data management, standard data analysis,
graphics, and "new statistics" (survey analysis, missing imputation,
generalized linear models, SEM, multilevel modeling ...). Except for
optimal scaling methods and some very specific ANOVA models, there is no
analysis offered by SPSS which I couldn't do using Stata. But more and
more I encounter situations where I have to tell colleagues that to the
best of my knowledge SPSS does not offer what I am able to do using Stata.

It took me about a month to become accustomed and about half a year to
become as efficient as I have been using SPSS. This investment in time
pays off: Using Stata improved the way I approach statistical problems
considerably. To my mind this is the most convincing reason for using Stata.

Since 2 years I am also using it in classes as an alternative to SPSS
(finally I made it mandatory). My experiences are very positive: Stata
can be used more efficiently to demonstrate basic statistic concepts, at
the same time it allows students to use advanced statistics with less
effort. Experience with using Stata in the computer lab shows that
students start feeling comfortable using syntax much faster as compared
to SPSS - perhaps due to Stata's excellent help system and because the
structure of Stata commands is simpler and more consistent than the
structure of SPSS commands.

Dirk

03-May-2013 19:28, John F Hall <[hidden email]> wrote:
 >
--- (snip) ---
>
> Has anyone actually displayed SPSS and Stata syntax side-by-side for the
> same basic jobs, data input, data dictionary, transformations,
> tabulation and graphics?
 >
--- (snip) ---


========================================
Dr. Dirk Enzmann
Institute of Criminal Sciences
Dept. of Criminology
Rothenbaumchaussee 33
D-20148 Hamburg
Germany

phone: [phone] (office)
        +49-(0)40-42838.4591 (Mrs Billon)
fax:   [phone]
email: [hidden email]
http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/Enzmann.html
========================================

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Enzmann]

Sorry the link to answer John's question should have been:

http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm

Dirk

Am 05.05.2013 14:47, schrieb Dirk Enzmann:

> Here it is:
> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0310&L=spssx-l&P=R14717
>
> BTW: "STATA" is no acronym and should be spelled "Stata".
>
> I used SPSS extensively for about 20 years. For a while I experimented
> with R to solve problems that could not be solved using SPSS, but I gave
> up - not completely, though - because R is far less user friendly than
> SPSS or Stata. Finally I made a move to Stata. The last 5 years I am
> using it almost exclusively and I am happy that I decided to make the
> move: Not only it is less expensive, at the same time it offers far more
> and better capabilities of data management, standard data analysis,
> graphics, and "new statistics" (survey analysis, missing imputation,
> generalized linear models, SEM, multilevel modeling ...). Except for
> optimal scaling methods and some very specific ANOVA models, there is no
> analysis offered by SPSS which I couldn't do using Stata. But more and
> more I encounter situations where I have to tell colleagues that to the
> best of my knowledge SPSS does not offer what I am able to do using Stata.
>
> It took me about a month to become accustomed and about half a year to
> become as efficient as I have been using SPSS. This investment in time
> pays off: Using Stata improved the way I approach statistical problems
> considerably. To my mind this is the most convincing reason for using
> Stata.
>
> Since 2 years I am also using it in classes as an alternative to SPSS
> (finally I made it mandatory). My experiences are very positive: Stata
> can be used more efficiently to demonstrate basic statistic concepts, at
> the same time it allows students to use advanced statistics with less
> effort. Experience with using Stata in the computer lab shows that
> students start feeling comfortable using syntax much faster as compared
> to SPSS - perhaps due to Stata's excellent help system and because the
> structure of Stata commands is simpler and more consistent than the
> structure of SPSS commands.
>
> Dirk
>
> 03-May-2013 19:28, John F Hall <[hidden email]> wrote:
>  >
> --- (snip) ---
>>
>> Has anyone actually displayed SPSS and Stata syntax side-by-side for the
>> same basic jobs, data input, data dictionary, transformations,
>> tabulation and graphics?
>  >
> --- (snip) ---

========================================
Dr. Dirk Enzmann
Institute of Criminal Sciences
Dept. of Criminology
Rothenbaumchaussee 33
D-20148 Hamburg
Germany

phone: [phone] (office)
        +49-(0)40-42838.4591 (Mrs Billon)
fax:   [phone]
email: [hidden email]
http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/Enzmann.html
========================================

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[John F Hall]

Dirk

This looks interesting.

I'll post some SPSS syntax for simple jobs on data from my site and see what
the Stata syntax is for the same job.  Only out of interest, mind: at my
age, I'm sticking to SPSS anyway.

John

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Dirk Enzmann
Sent: 05 May 2013 14:53
To: [hidden email]
Subject: Re: STATA

Sorry the link to answer John's question should have been:

http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm

Dirk

Am 05.05.2013 14:47, schrieb Dirk Enzmann:

> Here it is:
> http://www.listserv.uga.edu/cgi-bin/wa?A2=ind0310&L=spssx-l&P=R14717
>
> BTW: "STATA" is no acronym and should be spelled "Stata".
>
> I used SPSS extensively for about 20 years. For a while I experimented
> with R to solve problems that could not be solved using SPSS, but I
> gave up - not completely, though - because R is far less user friendly
> than SPSS or Stata. Finally I made a move to Stata. The last 5 years I
> am using it almost exclusively and I am happy that I decided to make
> the
> move: Not only it is less expensive, at the same time it offers far
> more and better capabilities of data management, standard data
> analysis, graphics, and "new statistics" (survey analysis, missing
> imputation, generalized linear models, SEM, multilevel modeling ...).
> Except for optimal scaling methods and some very specific ANOVA
> models, there is no analysis offered by SPSS which I couldn't do using
> Stata. But more and more I encounter situations where I have to tell
> colleagues that to the best of my knowledge SPSS does not offer what I am

able to do using Stata.

>
> It took me about a month to become accustomed and about half a year to
> become as efficient as I have been using SPSS. This investment in time
> pays off: Using Stata improved the way I approach statistical problems
> considerably. To my mind this is the most convincing reason for using
> Stata.
>
> Since 2 years I am also using it in classes as an alternative to SPSS
> (finally I made it mandatory). My experiences are very positive: Stata
> can be used more efficiently to demonstrate basic statistic concepts,
> at the same time it allows students to use advanced statistics with
> less effort. Experience with using Stata in the computer lab shows
> that students start feeling comfortable using syntax much faster as
> compared to SPSS - perhaps due to Stata's excellent help system and
> because the structure of Stata commands is simpler and more consistent
> than the structure of SPSS commands.
>
> Dirk
>
> 03-May-2013 19:28, John F Hall <[hidden email]> wrote:
>  >
> --- (snip) ---
>>
>> Has anyone actually displayed SPSS and Stata syntax side-by-side for
>> the same basic jobs, data input, data dictionary, transformations,
>> tabulation and graphics?
>  >
> --- (snip) ---

========================================
Dr. Dirk Enzmann
Institute of Criminal Sciences
Dept. of Criminology
Rothenbaumchaussee 33
D-20148 Hamburg
Germany

phone: [phone] (office)
        +49-(0)40-42838.4591 (Mrs Billon)
fax:   [phone]
email: [hidden email]
http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/En
zmann.html
========================================

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[John F Hall]

Stata syntax seems straightforward enough, but the output (from the hyperlinks) looks as if it came straight from a 1960s typewriter or a 1970s line-printer!

 

Social scientists will be more interested in the GSS than in data on 1978 cars.

 

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

 

 

 

 

 

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Dirk Enzmann
Sent: 05 May 2013 14:53
To: [hidden email]
Subject: Re: STATA

 

Sorry the link to answer John's question should have been:

 

http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm

 

Dirk

 

[Garry Gelade]

Hi All

 

I am running SPSS 18 with R 2.8.1.  I’d like to use the RStudio IDE which requires R2.11.1 or greater.

 

Question: If I migrate to R2.11.1, will it still work with SPSS 18?  And conversely, what version(s) of SPSS are compatible with R2.11.1?

 

Thanks for any advice

 

Garry

 

 

[Enzmann]

Well, layout is also a matter of taste (personally, I still prefer the "classical" output of SPSS 6.1 to fancy pivot tables, but this debate has been closed a long time ago ...)

As to the example: The auto dataset is famous because it is small, simple, but nevertheless useful for lots of examples as it contains various types of variables. Many Stata users know it by heart.

If you are looking for GSS or more advanced real data sets, see:
http://www.ats.ucla.edu/stat/stata/faq/sample_survey_setups.htm

Dirk

John F Hall wrote

Stata syntax seems straightforward enough, but the output (from the
hyperlinks) looks as if it came straight from a 1960s typewriter or a 1970s
line-printer!

Social scientists will be more interested in the GSS than in data on 1978
cars.


John F Hall (Mr)
[Retired academic survey researcher]

Email:    <mailto:[hidden email]> [hidden email]
Website:  <http://www.surveyresearch.weebly.com>
www.surveyresearch.weebly.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Dirk Enzmann
Sent: 05 May 2013 14:53
To: [hidden email]
Subject: Re: STATA

Sorry the link to answer John's question should have been:

 <http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm>
http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm

Dirk

[Jon K Peck]

Statistics versions are bound to a specific R version.  The source code and build instructions for the R plugins for Statistics is provided along with the compiled plugin, so you could build a plugin for a different R version.  In  most cases, this would amount to just substituting the appropriate R files used by the plugin and recompiling, but it is possible that code changes would be required.

Since you can have multiple R versions installed, it would be possible to use a tool like Rstudio for parts of the work, but, of course, the Statistics apis would not be available in that mode.

Statistics 21 - R2.14
Statistics 20 - R2.12
Statistics 19 - R2.10
Statistics 18 - R2.8


Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: [phone]




From:        Garry Gelade <[hidden email]>
To:        [hidden email],
Date:        05/06/2013 04:06 AM
Subject:        [SPSSX-L] R Versions
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Hi All
 
I am running SPSS 18 with R 2.8.1.  I’d like to use the RStudio IDE which requires R2.11.1 or greater.
 
Question: If I migrate to R2.11.1, will it still work with SPSS 18?  And conversely, what version(s) of SPSS are compatible with R2.11.1?
 
Thanks for any advice
 
Garry
 
 

[Dale]

If it matters I regularly use data files in the 100,000 cases range with 200-300 variables.  On a 64 bit system with 32 gig running an i7 at 3.40 for clock speed it take some time with the large files but generally does not require much of most operations with 10,000 files and 200-300 variables in the data set.

Dale

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: Friday, May 03, 2013 8:14 AM
To: [hidden email]
Subject: Re: STATA

For several years I have been using SYSTAT as my base of operations with SPSS as one of the supplemental.  For what I need most of the time it is just a better alternative.   There are things SPSS currently does very well relative to SYSTAT, but the need to add-on modules, rental rather than own type licensing, and other issues have just not been a positive for me.  It may be something to look into as well.

Dale


Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment University of South Dakota Slagle Hall Room 102
414 East Clark Street
605-677-6497



-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Stan Gorodenski
Sent: Friday, May 03, 2013 1:23 AM
To: [hidden email]
Subject: STATA

I have looked into STATA as an alternative to SPSS. The statistics it has appears to be every bit as good as SPSS's. Even its ARIMA procedure is more flexible and more powerful like the original PACK program than SPSS. For years SPSS has not brought up its ARIMA procedure to where it should be, and this still hasn't happened under IBM (I don't think). Its programming capabilities, although it would require learning a lot of new syntax, appear to be just as good and maybe even better than SPSS's.
Does anyone have any experience with STATA to verify this and to say where STATA might fall short compared to SPSS, and be a disappointment?
Stan

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[John F Hall]

I've got plenty of data sets of my own, from major national surveys to the
small (cumulative) data set generated by students on my courses.  See:
http://surveyresearch.weebly.com/data-sets-used-in-survey-analysis-workshop.
html

I don't have the time, but you're welcome to replicate the exercises in my
tutorials using Stata instead of SPSS, then display the syntax and tables
side by side.

Many years ago, Beverley Rowe and I had the idea of organising a (knockout?)
competition to see who could take a raw data set from a questionnaire survey
and turn it into a fully documented system file in the shortest possible
time using their favourite software.

Any takers?

John Hall

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Enzmann
Sent: 06 May 2013 13:01
To: [hidden email]
Subject: Re: STATA

Well, layout is also a matter of taste (personally, I still prefer the
"classical" output of SPSS 6.1 to fancy pivot tables, but this debate has
been closed a long time ago ...)

As to the example: The auto dataset is famous because it is small, simple,
but nevertheless useful for lots of examples as it contains various types of
variables. Many Stata users know it by heart.

If you are looking for GSS or more advanced real data sets, see:
http://www.ats.ucla.edu/stat/stata/faq/sample_survey_setups.htm

Dirk


John F Hall wrote

> Stata syntax seems straightforward enough, but the output (from the
> hyperlinks) looks as if it came straight from a 1960s typewriter or a
> 1970s line-printer!
>
> Social scientists will be more interested in the GSS than in data on
> 1978 cars.
>
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email:    &lt;mailto:

> johnfhall@

> &gt;

> johnfhall@

> Website:  &lt;http://www.surveyresearch.weebly.com&gt;
> www.surveyresearch.weebly.com
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:

> SPSSX-L@.UGA

> ] On Behalf Of
> Dirk Enzmann
> Sent: 05 May 2013 14:53
> To:

> SPSSX-L@.UGA

> Subject: Re: STATA
>
> Sorry the link to answer John's question should have been:
>
>
> &lt;http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm&g
> t; http://www.ats.ucla.edu/stat/stata/faq/spss_command_to_stata.htm
>
> Dirk





--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/STATA-tp5719891p5719994.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD