SPSSX Discussion

SPSS Statistics Survey

Classic

List

Threaded

24 messages Options

John F Hall

SPSS Statistics Survey

The survey SPSS are doing https://s.userzoom.com/m/MSBDOVM5NDUg breaks a few cardinal rules in questionnaire design. It contains double questions, check-lists which are insufficiently specific, allows no button for genuine “Never use” or “Really can’t say” responses, and won’t allow progress unless every item is answered, thus eliciting spurious (mid-point or wish-list) responses in order to continue.

Biggest problem is you can’t skip backwards and forwards or review your responses before they are finally submitted. I finished up copy/pasting the questions and my answers and also MS-Snipping the graphics and list sections page by page in order to have a record.

Ironic since SPSS originated in survey research, but the survey could come in useful to show students and beginning researchers how not to design a questionnaire.

The high cost of SPSS means academic institutions are turning in their hundreds to Stata, which has nothing like as good data management facilities or output.

The last question is:

Please give us any final comments, wish list items, or suggestions. You can also use this space to comment on this survey or other aspects of your experience with IBM SPSS Statistics.

To which I have replied (in addition to earlier wish-lists):

You really have to get the cost down for academic users: you are going to lose out to Stata and then a whole future generation of SPSS clients. SPSS has no equal in data management. The origins of SPSS are based in survey research: perhaps you should start learning how to do a survey and offer SPSS to academic institutions as a not-for-profit service, not a way to make money.

But will they listen?

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

Christopher Stride

Re: SPSS Statistics Survey

Whilst I agree with your sentiments re: their need to cut costs so as to not lose the 'seeding' grounds of Universities, I suspect they'd be more likely to listen to a more polite response than "perhaps you should start learning how to do a survey..."

On 28/01/2015 15:09, John F Hall wrote:

The survey SPSS are doing https://s.userzoom.com/m/MSBDOVM5NDUg breaks a few cardinal rules in questionnaire design. It contains double questions, check-lists which are insufficiently specific, allows no button for genuine “Never use” or “Really can’t say” responses, and won’t allow progress unless every item is answered, thus eliciting spurious (mid-point or wish-list) responses in order to continue.

Biggest problem is you can’t skip backwards and forwards or review your responses before they are finally submitted. I finished up copy/pasting the questions and my answers and also MS-Snipping the graphics and list sections page by page in order to have a record.

Ironic since SPSS originated in survey research, but the survey could come in useful to show students and beginning researchers how not to design a questionnaire.

The high cost of SPSS means academic institutions are turning in their hundreds to Stata, which has nothing like as good data management facilities or output.

The last question is:

Please give us any final comments, wish list items, or suggestions. You can also use this space to comment on this survey or other aspects of your experience with IBM SPSS Statistics.

To which I have replied (in addition to earlier wish-lists):

You really have to get the cost down for academic users: you are going to lose out to Stata and then a whole future generation of SPSS clients. SPSS has no equal in data management. The origins of SPSS are based in survey research: perhaps you should start learning how to do a survey and offer SPSS to academic institutions as a not-for-profit service, not a way to make money.

But will they listen?

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== 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

Martin Holt-3

Re: SPSS Statistics Survey

In reply to this post by John F Hall

Dear John~

Another major player in this field is SAS. How would you rate SPSS vs. SAS?

Regards,

Martin

Martin P. Holt

Freelance Medical Statistician and Quality Expert

[hidden email]

Persistence and Determination Alone are Omnipotent !

If you can't explain it simply, you don't understand it well enough.....Einstein

Linked In: https://www.linkedin.com/profile/edit?trk=nav_responsive_sub_nav_edit_profile

From: John F Hall <[hidden email]>
To: [hidden email]
Sent: Wednesday, 28 January 2015, 15:09
Subject: SPSS Statistics Survey

The survey SPSS are doing https://s.userzoom.com/m/MSBDOVM5NDUg breaks a few cardinal rules in questionnaire design. It contains double questions, check-lists which are insufficiently specific, allows no button for genuine “Never use” or “Really can’t say” responses, and won’t allow progress unless every item is answered, thus eliciting spurious (mid-point or wish-list) responses in order to continue.

Biggest problem is you can’t skip backwards and forwards or review your responses before they are finally submitted. I finished up copy/pasting the questions and my answers and also MS-Snipping the graphics and list sections page by page in order to have a record.

Ironic since SPSS originated in survey research, but the survey could come in useful to show students and beginning researchers how not to design a questionnaire.

The high cost of SPSS means academic institutions are turning in their hundreds to Stata, which has nothing like as good data management facilities or output.

The last question is:

Please give us any final comments, wish list items, or suggestions. You can also use this space to comment on this survey or other aspects of your experience with IBM SPSS Statistics.

To which I have replied (in addition to earlier wish-lists):

You really have to get the cost down for academic users: you are going to lose out to Stata and then a whole future generation of SPSS clients. SPSS has no equal in data management. The origins of SPSS are based in survey research: perhaps you should start learning how to do a survey and offer SPSS to academic institutions as a not-for-profit service, not a way to make money.

But will they listen?

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

Email: [hidden email]
Website: www.surveyresearch.weebly.com
SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== 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

Re: SPSS Statistics Survey

In reply to this post by John F Hall

Bob, Martin, Chris, Tim

[and and others who may respond]

If you can teach R to reluctant 2^nd year sociology undergraduates in 13 x 1 hour classroom sessions and 13 x 2-hour lab sessions, sufficient to enable them to do useful work on their own research projects and subsequently on professional placements or in their first research posts, I might agree with you.

I don’t have access to Stata, but there are some comparisons of Stata and SPSS for one or two simple exercises from my tutorials, replicated by Dirk Enzmann and Jeff Meyer. The GUI looks reasonable but the tables looked like 30-year old line-printer output.

You can see these on my page, http://surveyresearch.weebly.com/stata-and-spss.html but there’s not much on it yet.

My teaching of social research and survey methods dates from 1968, and my use of SPSS from 1972, hence my adherence to SPSS. I’m a (very experienced) survey researcher with a sociological tinge, not a statistician, and I suspect many of the comments received are from a statistics perspective. I have no experience of SAS although I am aware of it. I tried R once and ran. I have been (early) retired for almost 23 years, have no affiliation to any institution, and no resources other than my own time and good-will, so cannot afford to purchase or learn any other software to a level where I would be comfortable teaching it. If anyone fancies writing parallel tutorials (in SAS, Stata or R) to any or all of those on my site, I’d welcome you as guest author(s).

Now into my 75^th year, I need to concentrate on completing the planned SPSS tutorials on my site before I pop my clogs, whilst continuing to explore and revise SPSS files for major surveys in the public domain, many of which need serious work to make them easier to use by teachers, students and researchers.

John

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

From: Gerzoff, Bob (CDC/ONDIEH/NCCDPHP) [mailto:[hidden email]]
Sent: 28 January 2015 16:52
To: John F Hall
Subject: RE: SPSS Statistics Survey

Lose out to STATA? R is free and even SAS is free now for academic use.

From: Survey Research Methods Section of the ASA [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 28, 2015 10:10 AM
To: [hidden email]
Subject: SPSS Statistics Survey

Ironic since SPSS originated in survey research, but the survey could come in useful to show students and beginning researchers how not to design a questionnaire.

The high cost of SPSS means academic institutions are turning in their hundreds to Stata, which has nothing like as good data management facilities or output.

The last question is:

Please give us any final comments, wish list items, or suggestions. You can also use this space to comment on this survey or other aspects of your experience with IBM SPSS Statistics.

To which I have replied (in addition to earlier wish-lists):

But will they listen?

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ To subscribe/unsubscribe SRMSNet: http://listserv.umd.edu/cgi-bin/wa?A0=srmsnet&D=0&F=&H=0&O=T&S=&T=1

SRMS website: http://www.amstat.org/sections/srms/ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Bruce Weaver

Re: SPSS Statistics Survey

Administrator

Down in the bowels of John F Hall's post, Bob Gerzoff said:

"R is free and even SAS is free now for academic use."

I'd not heard that before. I assume Bob is referring to "SAS OnDemand for Academics".

http://www.sas.com/en_us/industry/higher-education/on-demand-for-academics.html

John F Hall wrote

Bob, Martin, Chris, Tim

[and and others who may respond]

If you can teach R to reluctant 2nd year sociology undergraduates in 13 x 1
hour classroom sessions and 13 x 2-hour lab sessions, sufficient to enable
them to do useful work on their own research projects and subsequently on
professional placements or in their first research posts, I might agree with
you.

I don't have access to Stata, but there are some comparisons of Stata and
SPSS for one or two simple exercises from my tutorials, replicated by Dirk
Enzmann and Jeff Meyer. The GUI looks reasonable but the tables looked like
30-year old line-printer output.

You can see these on my page,
http://surveyresearch.weebly.com/stata-and-spss.html but there's not much
on it yet.
My teaching of social research and survey methods dates from 1968, and my
use of SPSS from 1972, hence my adherence to SPSS. I'm a (very experienced)
survey researcher with a sociological tinge, not a statistician, and I
suspect many of the comments received are from a statistics perspective. I
have no experience of SAS although I am aware of it. I tried R once and
ran. I have been (early) retired for almost 23 years, have no affiliation
to any institution, and no resources other than my own time and good-will,
so cannot afford to purchase or learn any other software to a level where I
would be comfortable teaching it. If anyone fancies writing parallel
tutorials (in SAS, Stata or R) to any or all of those on my site, I'd
welcome you as guest author(s).

Now into my 75th year, I need to concentrate on completing the planned SPSS
tutorials on my site before I pop my clogs, whilst continuing to explore and
revise SPSS files for major surveys in the public domain, many of which need
serious work to make them easier to use by teachers, students and
researchers.

John

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

Email: <mailto:[hidden email]> [hidden email]
Website: <http://www.surveyresearch.weebly.com/>
www.surveyresearch.weebly.com
SPSS start page:
<http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
www.surveyresearch.weebly.com/1-survey-analysis-workshop

From: Gerzoff, Bob (CDC/ONDIEH/NCCDPHP) [mailto:[hidden email]]
Sent: 28 January 2015 16:52
To: John F Hall
Subject: RE: SPSS Statistics Survey

Lose out to STATA? R is free and even SAS is free now for academic use.

From: Survey Research Methods Section of the ASA
[mailto:[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 28, 2015 10:10 AM
To: [hidden email]
Subject: SPSS Statistics Survey

The survey SPSS are doing <https://s.userzoom.com/m/MSBDOVM5NDUg>
https://s.userzoom.com/m/MSBDOVM5NDUg breaks a few cardinal rules in
questionnaire design. It contains double questions, check-lists which are
insufficiently specific, allows no button for genuine "Never use" or "Really
can't say" responses, and won't allow progress unless every item is
answered, thus eliciting spurious (mid-point or wish-list) responses in
order to continue.

Biggest problem is you can't skip backwards and forwards or review your
responses before they are finally submitted. I finished up copy/pasting the
questions and my answers and also MS-Snipping the graphics and list sections
page by page in order to have a record.

Ironic since SPSS originated in survey research, but the survey could come
in useful to show students and beginning researchers how not to design a
questionnaire.

The high cost of SPSS means academic institutions are turning in their
hundreds to Stata, which has nothing like as good data management facilities
or output.

The last question is:

Please give us any final comments, wish list items, or suggestions. You can
also use this space to comment on this survey or other aspects of your
experience with IBM SPSS Statistics.

To which I have replied (in addition to earlier wish-lists):

You really have to get the cost down for academic users: you are going to
lose out to Stata and then a whole future generation of SPSS clients. SPSS
has no equal in data management. The origins of SPSS are based in survey
research: perhaps you should start learning how to do a survey and offer
SPSS to academic institutions as a not-for-profit service, not a way to make
money.

But will they listen?

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

Email: [hidden email]
Website: www.surveyresearch.weebly.com
<http://www.surveyresearch.weebly.com/>
SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop
<http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ To subscribe/unsubscribe SRMSNet:
http://listserv.umd.edu/cgi-bin/wa?A0=srmsnet
<http://listserv.umd.edu/cgi-bin/wa?A0=srmsnet&D=0&F=&H=0&O=T&S=&T=1>
&D=0&F=&H=0&O=T&S=&T=1
SRMS website: http://www.amstat.org/sections/srms/
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

=====================
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

--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

PLEASE NOTE THE FOLLOWING:
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/).

John F Hall

Re: SPSS Statistics Survey

Bruce

That looks like a very interesting free resource. I'll check it out fully
tomorrow.

John

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bruce Weaver
Sent: 28 January 2015 19:57
To: [hidden email]
Subject: Re: SPSS Statistics Survey

Down in the bowels of John F Hall's post, Bob Gerzoff said:

"R is free and even SAS is free now for academic use."

I'd not heard that before. I assume Bob is referring to "SAS OnDemand for
Academics".

http://www.sas.com/en_us/industry/higher-education/on-demand-for-academics.h
tml

John F Hall wrote

> Bob, Martin, Chris, Tim
>
> [and and others who may respond]
>
> If you can teach R to reluctant 2nd year sociology undergraduates in
> 13 x
> 1
> hour classroom sessions and 13 x 2-hour lab sessions, sufficient to
> enable them to do useful work on their own research projects and
> subsequently on professional placements or in their first research
> posts, I might agree with you.
>
> I don't have access to Stata, but there are some comparisons of Stata
> and SPSS for one or two simple exercises from my tutorials, replicated
> by Dirk Enzmann and Jeff Meyer. The GUI looks reasonable but the
> tables looked like 30-year old line-printer output.
>
> You can see these on my page,
> http://surveyresearch.weebly.com/stata-and-spss.html but there's not
> much on it yet.
> My teaching of social research and survey methods dates from 1968, and
> my use of SPSS from 1972, hence my adherence to SPSS. I'm a (very
> experienced)
> survey researcher with a sociological tinge, not a statistician, and I
> suspect many of the comments received are from a statistics perspective.
> I
> have no experience of SAS although I am aware of it. I tried R once
> and ran. I have been (early) retired for almost 23 years, have no
> affiliation to any institution, and no resources other than my own
> time and good-will, so cannot afford to purchase or learn any other
> software to a level where I would be comfortable teaching it. If
> anyone fancies writing parallel tutorials (in SAS, Stata or R) to any
> or all of those on my site, I'd welcome you as guest author(s).
>
> Now into my 75th year, I need to concentrate on completing the planned
> SPSS tutorials on my site before I pop my clogs, whilst continuing to
> explore and revise SPSS files for major surveys in the public domain,
> many of which need serious work to make them easier to use by
> teachers, students and researchers.
>
> John
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email: <mailto:

> johnfhall@

> >

> johnfhall@

>
> Website: <http://www.surveyresearch.weebly.com/>
> www.surveyresearch.weebly.com
> SPSS start page:
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html&g
> t; www.surveyresearch.weebly.com/1-survey-analysis-workshop
>
>
>
>
> From: Gerzoff, Bob (CDC/ONDIEH/NCCDPHP) [mailto:

> rcg8@

> ]
> Sent: 28 January 2015 16:52
> To: John F Hall
> Subject: RE: SPSS Statistics Survey
>
> Lose out to STATA? R is free and even SAS is free now for academic use.
>
> From: Survey Research Methods Section of the ASA
> [mailto:

> SRMSNET@.UMD

> ] On Behalf Of John F Hall
> Sent: Wednesday, January 28, 2015 10:10 AM
> To:

> SRMSNET@.UMD

> Subject: SPSS Statistics Survey
>
> The survey SPSS are doing
> <https://s.userzoom.com/m/MSBDOVM5NDUg>
> https://s.userzoom.com/m/MSBDOVM5NDUg breaks a few cardinal rules in
> questionnaire design. It contains double questions, check-lists which
> are insufficiently specific, allows no button for genuine "Never use"
> or "Really can't say" responses, and won't allow progress unless every
> item is answered, thus eliciting spurious (mid-point or wish-list)
> responses in order to continue.
>
> Biggest problem is you can't skip backwards and forwards or review
> your responses before they are finally submitted. I finished up
> copy/pasting the questions and my answers and also MS-Snipping the
> graphics and list sections page by page in order to have a record.
>
> Ironic since SPSS originated in survey research, but the survey could
> come in useful to show students and beginning researchers how not to
> design a questionnaire.
>
> The high cost of SPSS means academic institutions are turning in their
> hundreds to Stata, which has nothing like as good data management
> facilities or output.
>
> The last question is:
>
> Please give us any final comments, wish list items, or suggestions.
> You can also use this space to comment on this survey or other aspects
> of your experience with IBM SPSS Statistics.
>
> To which I have replied (in addition to earlier wish-lists):
>
> You really have to get the cost down for academic users: you are going
> to lose out to Stata and then a whole future generation of SPSS clients.
> SPSS
> has no equal in data management. The origins of SPSS are based in
> survey
> research: perhaps you should start learning how to do a survey and
> offer SPSS to academic institutions as a not-for-profit service, not a
> way to make money.
>
> But will they listen?
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email:

> johnfhall@

>
> Website: www.surveyresearch.weebly.com
> <http://www.surveyresearch.weebly.com/>
> SPSS start page:
> www.surveyresearch.weebly.com/1-survey-analysis-workshop
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html&g
> t;
>
>
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ To subscribe/unsubscribe SRMSNet:
> http://listserv.umd.edu/cgi-bin/wa?A0=srmsnet
> <http://listserv.umd.edu/cgi-bin/wa?A0=srmsnet&D=0&F=&H
> =0&O=T&S=&T=1>
> &D=0&F=&H=0&O=T&S=&T=1
> SRMS website: http://www.amstat.org/sections/srms/
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

> (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

-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/SPSS-Statistics-Survey-tp57285
13p5728522.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

John F Hall

FW: SPSS Statistics Survey

In reply to this post by John F Hall

From: Machteld Hoskens [[hidden email]]
Sent: 29 January 2015 17:23
To: John F Hall
Subject: SPSS Statistics Survey

Yes, go ahead!

machteld

From: John F Hall [[hidden email]]
Sent: donderdag 29 januari 2015 17:20
To: Machteld Hoskens
Subject: RE: SPSS or what?

Machteld

This really should be forwarded to the SPSS and QM lists. Would you like me to do it for you?

John

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

From: Machteld Hoskens [[hidden email]]
Sent: 29 January 2015 17:07
To: [hidden email]
Subject: FW: SPSS or what?

In reply to your question: ‘Does anyone know current prices for SPSS? I can’t seem to find any on the IBM website. As for impact, please share your experience.’ on the quantitative methods teaching forum:

As you know, UCLA is an important source for ins and outs of statistics and statistical software, so especially here it would be useful for IBM to make efforts to please and keep future SPSS users…

machteld

From: JANE TOROUS [[hidden email]]
Sent: maandag 7 oktober 2013 19:52
Subject: Re: SPSS or what?

I appreciate this dialogue about SPSS, as my recent transactions with the company suggest it is time to find a new program.

Their support for graduate students is nil.

I contacted SPSS support on behalf of students who work with SPSS Gradpack.

We had encountered a bug in the chart dialogs. We reported this to

SPSS/IBM but they were unwilling to follow up because they don't support Gradpack.

Only as a "favor" did SPSS/IBM investigate the matter. They suggested different fixes, which did not work. When they realized that it might be a bug in the chart code, they ceased to correspond with us.

How can a company improve its product, if it shuns customer feedback and information...particularly academic ones?

Jane Gould, PhD

[hidden email]

On Mon, Oct 7, 2013 at 10:45 AM, JANE TOROUS <[hidden email]> wrote:

I appreciate this dialogue on SPSS, as recent transactions with the company have made me think it is also time to find a new package.

I contacted SPSS support on behalf of students who work with SPSS Gradpack (it is an annual license renewal).

We had encountered a bug in the chart dialogs. We reported this to

SPSS/IBM but they were unwilling to follow up because the license was for

Gradpack and they didn't support that.

Only as a "favor" did SPSS/IBM investigate the matter, and when they realized that it might be a bug

in the chart dialogs, they ceased correspondence. The staff seems to be located far-away, and perhaps off shore.

How can a company improve its product, if it shuns customer feedback and information?

Jane Gould, PhD

[hidden email]

On Mon, Oct 7, 2013 at 10:00 AM, J. Ann Selzer <[hidden email]> wrote:

This thread brings back a lot of memories. When I was in grad school in the late 70s/early 80s, it was observed that the communication researchers (as opposed to the rhetoricians in the department) tended to work on projects in teams. What we knew that they didn't was it took more than one person to carry the boxes of computer cards, the SPSS and BMDP manuals to and from the computer center. Not the Stone Age, exactly. More like the Card Age. JAS

J. Ann Selzer, Ph.D
Selzer & Company
Des Moines, Iowa 50309

For purposes of this list, use [hidden email]
For other purposes, use [hidden email]

-----Original Message-----
From: Rich, Chuck <[hidden email]>
To: AAPORnet <[hidden email]>
Sent: Fri, Oct 4, 2013 6:53 am
Subject: RE: [aapor-net] SPSS or what?

All,

I first used SAS in grad school in 1973 or 74, shortly after it was made available. I have used SAS continuously since the late 70’s, and have also used BMDP (mainframe), and SPSS (also mostly on the mainframe). Some of you may remember the old MULTIVARIANCE program using punch cards. While the discussion here focuses mainly on statistics, I just want to add that I consider SAS to be much superior in terms of programming flexibility. This would include data manipulation, SQL-types of queries, merges, joins, transposing data sets, using arrays and loops, macro facilities in which entire procedures can be invoked iteratively, creation of various types of output files in many formats (using the ODS facility), using functions, being able to structure your code in any way that seems appropriate (rather than being confined to point-and-click format or other rigid formats), and so on—basically programming tasks rather than statistical ones.

I can’t speak with authority about how SAS compares to the other packages in terms of statistics, though I think it would be very comparable in quality. But as far as teaching statistics goes, and relatively small consulting operations, probably SPSS and some of these other packages would be more practical. SAS would be very expensive unless you are working under a site license.

Best,

Chuck

From: Jason Husser [[hidden email]]
Sent: Thursday, October 03, 2013 7:27 PM
To: AAPORnet
Cc: AAPORnet
Subject: Re: [aapor-net] SPSS or what?

I teach political science research methods and survey methods to undergraduates at Elon. I tell my undergraduates something similar to Woody. I also tell them that some jobs of potential interest (especially those working with private medical records) may require SAS.

However, many students (particularly 19 year old sophomores) have little idea about the direction of their future careers. So, I start with Excel because it is almost always familiar to them. I quickly transition and spend most of the semester with SPSS. I intermittently return them to Excel to work on making tables and figures for reports. I will soon offer about a month of exposure to Stata and introduce them to R for at least a class session. To ensure that they've at least heard of other relevant systems, I also spend some time to provide a very brief discussion/demonstration of SQL, PHP, Python, interactive web graphics, and the clunky outputs from our CATI software.

This approach of breadth over depth isn't necessarily the ideal way, but I find it offers a broad foundation for future learning and gets them adequately prepared to conduct analysis for good undergraduate research papers and, hopefully, basic career and internship tasks. Other faculty members and I are readily available to help them outside of class if they need deeper instruction in a particular software. My approach might not work if I didn't have small classes, a dedicated computer lab classroom, and 1 hour 40 minute class sessions.

-Jason

Jason Husser, Ph.D.
Assistant Professor of Political Science
Assistant Director of the Elon University Poll
Elon University
[hidden email]

<a href="tel:%28985%29%20516-7830" target="_blank">(985) 516-7830

On Thu, Oct 3, 2013 at 6:38 PM, Losh, Susan <[hidden email]> wrote:

As someone who teaches loglinear analysis and multinomial logistic regression all I can say is OUCH!

I try to take one day at a time but sometimes several days attack me all at once. Anonymous.
Email is the fastest way to reach me.

Susan Carol Losh, PhD
Department of Educational Psychology and Learning Systems
Florida State University
Tallahassee FL 32306-4453

850.644.8778 VOICE

850.644.8776 FAX

Editor: Bulletin of Science Technology & Society
Chair, AERA Special Interest Group: Advanced Studies of National Databases
American Statistical Association/NSF Research Fellow
http://mailer.fsu.edu/~slosh/index.html
________________________________________

From: Milton R. Goldsamt [[hidden email]]

Sent: Thursday, October 03, 2013 5:56 PM
To: AAPORnet
Subject: Re: [aapor-net] SPSS or what?

In my professional career, I’ve used BMD (remember that?), Excel, SPSS and SAS. Now retired, consulting some, and extensively tutoring college/grad students from online and local Washington area universities, I find that SAS is rarely used (one such student needed it in my last three years of tutoring). I also find that most consulting firms I deal with use SPSS (on which it’s easy to train their staffs), and students I’ve run into are taking stat courses that turn to SPSS for computing almost all of their analyses. (I now have one student using Qualtrics and SPSS for his dissertation.)

However, these students are also exposed for the most part to a Gradpack version of SPSS that contains a wider variety of procedures than the Base version 21 of SPSS, now owned by IBM. Apparently IBM SPSS wants to expose them to a wide range of procedures (such as logistic regression, Repeated Measures, MANOVA, etc.) and then have them buy multiple modules, not just Base SPSS, when they leave the academic world.

Thus SPSS can be expensive if your computing needs are wide ones.

Milton Goldsamt

Milton R. Goldsamt, Ph.D.
Consulting Research Psychologist & Statistician
Silver Spring, MD
[hidden email]

301-649-2768
(C) 240-671-7201

On Oct 3, 2013, at 4:29 PM, Phillip Downs <[hidden email]> wrote:

> Having taught at universities for 37 years and run my own business for 30, Woody's advice seems solid to me.
>
> Phillip Downs, Ph.D.
> Senior Partner | Kerr & Downs Research
> Founder | TallahasseeVoices
> Professor of Marketing | Florida State University, retired

> ph. 850-906-3111 | fax 850-906-3112

>
>
>
> -----Original Message-----
> From: Woody Carter [[hidden email]]
> Sent: Thursday, October 03, 2013 3:51 PM
> To: AAPORnet
> Subject: RE: [aapor-net] SPSS or what?
>
> I tell my students:
>
> If you plan to be a stat guru, master R.
>
> If you plan to go into business, master SPSS and Excel.
>
> If you plan to go into economics, poli sci, or soc, master Stata.
>
> If you plan to go into the nonprofit world, master Excel.
>
> Woody
>
> -----Original Message-----

> From: [hidden email] [[hidden email]]
> Sent: Thursday, October 03, 2013 2:32 PM
> To: AAPORnet
> Subject: [aapor-net] SPSS or what?
>
> This discussion has reminded me that I wanted to ask aapornetters about the future of SPSS in academia and ultimately business. I first started working with SPSS in the 1980s. I eventually also learned SAS. In my years in the private sector, SPSS and to some extent SAS were the packages I saw to be most used. Now that I'm in a university again, I find that the software used is often R (open source and free) or STATA.
>
> I still have v13 on my PC at home, but I'm about to replace this 9-year old XP. I'd like to give students marketable working skills, and I'd like to use what's compatible with my client needs, but I'm really dissatisfied with the latest versions of SPSS (e.g., ordinal regression has been
> removed) and my colleagues have raised other issues around the accuracy of standard errors.
>
> So I'm wondering...
> 1. For those of you in academia, what are your students using and why?
> 2. For those of you not in academia, what do you want to use, or want your employees to use?
> 3. And more specifically, which do both groups prefer - SPSS, SAS, STATA or R?
>
> Thanks!
> Leora
>
>
> ---------------------------------------------
> Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines
>
> Questions? Contact [hidden email]
>
> ---------------------------------------------
> Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines
>
> Questions? Contact [hidden email]

> -----
> No virus found in this message.
> Checked by AVG - www.avg.com

> Version: 2013.0.3408 / Virus Database: 3222/6710 - Release Date: 09/30/13
>
>
>
> ---------------------------------------------
> Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines
>
> Questions? Contact [hidden email]

---------------------------------------------

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

---------------------------------------------
Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

John F Hall

FW: SPSS Statistics Survey

In reply to this post by John F Hall

From: Machteld Hoskens [[hidden email]]
Sent: 29 January 2015 17:23
To: John F Hall
Subject: SPSS Statistics Survey

Yes, go ahead!

machteld

From: John F Hall [[hidden email]]
Sent: donderdag 29 januari 2015 17:20
To: Machteld Hoskens
Subject: RE: SPSS or what?

Machteld

This really should be forwarded to the SPSS and QM lists. Would you like me to do it for you?

John

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

From: Machteld Hoskens [[hidden email]]
Sent: 29 January 2015 17:07
To: [hidden email]
Subject: FW: SPSS or what?

As you know, UCLA is an important source for ins and outs of statistics and statistical software, so especially here it would be useful for IBM to make efforts to please and keep future SPSS users…

machteld

I appreciate this dialogue about SPSS, as my recent transactions with the company suggest it is time to find a new program.

Their support for graduate students is nil.

I contacted SPSS support on behalf of students who work with SPSS Gradpack.

We had encountered a bug in the chart dialogs. We reported this to

SPSS/IBM but they were unwilling to follow up because they don't support Gradpack.

How can a company improve its product, if it shuns customer feedback and information...particularly academic ones?

Jane Gould, PhD

[hidden email]

On Mon, Oct 7, 2013 at 10:45 AM, JANE TOROUS <[hidden email]> wrote:

I appreciate this dialogue on SPSS, as recent transactions with the company have made me think it is also time to find a new package.

I contacted SPSS support on behalf of students who work with SPSS Gradpack (it is an annual license renewal).

We had encountered a bug in the chart dialogs. We reported this to

SPSS/IBM but they were unwilling to follow up because the license was for

Gradpack and they didn't support that.

Only as a "favor" did SPSS/IBM investigate the matter, and when they realized that it might be a bug

in the chart dialogs, they ceased correspondence. The staff seems to be located far-away, and perhaps off shore.

How can a company improve its product, if it shuns customer feedback and information?

Jane Gould, PhD

[hidden email]

On Mon, Oct 7, 2013 at 10:00 AM, J. Ann Selzer <[hidden email]> wrote:

J. Ann Selzer, Ph.D
Selzer & Company
Des Moines, Iowa 50309

For purposes of this list, use [hidden email]
For other purposes, use [hidden email]

-----Original Message-----
From: Rich, Chuck <[hidden email]>
To: AAPORnet <[hidden email]>
Sent: Fri, Oct 4, 2013 6:53 am
Subject: RE: [aapor-net] SPSS or what?

All,

Best,

Chuck

From: Jason Husser [[hidden email]]
Sent: Thursday, October 03, 2013 7:27 PM
To: AAPORnet
Cc: AAPORnet
Subject: Re: [aapor-net] SPSS or what?

-Jason

Jason Husser, Ph.D.
Assistant Professor of Political Science
Assistant Director of the Elon University Poll
Elon University
[hidden email]

<a href="tel:%28985%29%20516-7830" target="_blank">(985) 516-7830

On Thu, Oct 3, 2013 at 6:38 PM, Losh, Susan <[hidden email]> wrote:

850.644.8778 VOICE

From: Milton R. Goldsamt [[hidden email]]

Sent: Thursday, October 03, 2013 5:56 PM
To: AAPORnet
Subject: Re: [aapor-net] SPSS or what?

301-649-2768
(C) 240-671-7201

On Oct 3, 2013, at 4:29 PM, Phillip Downs <[hidden email]> wrote:

> ph. 850-906-3111 | fax 850-906-3112

> -----
> No virus found in this message.
> Checked by AVG - www.avg.com

---------------------------------------------

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

Information about AAPORnet can be found in the AAPORnet Guidelines at: http://www.aapor.org/AAPORnetGuidelines

Questions? Contact [hidden email]

David Marso

Re: FW: SPSS Statistics Survey

Administrator

In reply to this post by John F Hall

Chuck wrote:
"I just want to add that I consider SAS to be much superior in terms of programming flexibility. This would include data manipulation, SQL-types of queries, merges, joins, transposing data sets, using arrays and loops, macro facilities in which entire procedures can be invoked iteratively, creation of various types of output files in many formats (using the ODS facility), using functions, being able to structure your code in any way that seems appropriate (rather than being confined to point-and-click format or other rigid formats), and so on—basically programming tasks rather than statistical ones."

I don't know what version of SPSS Chuck refers to but he does everyone disservice by being terribly misinformed and passing on this disinformation with respect to the capabilities of SPSS. I surmise pre version 4 (1990ish) or SPSSPC+ ???
SPSS has supported ALL OF THE ABOVE for as long as I can remember!!!
Please be informed before opening your pie hole?
SQL-types of queries ->
merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
using arrays and loops-> VECTOR, LOOP, DO REPEAT
macro facility-> DEFINE !ENDDEFINE
creation of various types of output files in many formats -> OMS, OUTPUT SAVE
Code can be structured in any way one sees fit.
In NO WAY re people confined to "Point and Click".
Yeah, Basically PROGRAMMING!!!!!!!
In fact, I challenge Chuck to come up with any task doable in SAS that can't be done (by someone like me) more elegantly using SPSS!!

Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"

Bruce Weaver

Re: FW: SPSS Statistics Survey

Administrator

In my experience, the chain of "reasoning" behind statements such as Chuck's is something like this:

1. I don't know how to perform some task using SPSS.
2. The novice, point-and-click-only SPSS users I know do not know how to do it either.
3. Therefore, SPSS cannot do it.
4. I do know how to perform the task in [fill in some other stats package here].
5. Therefore [the other stats package] is vastly superior to SPSS.

The question that is typically lacking, as suggested by David's post, is whether a more advanced SPSS user (not limited to point-and-click-only) knows how to perform the task.

p.s. - I hope some SAS users accept David's challenge. That would be fun to watch! ;-)

David Marso wrote

Chuck wrote:
"I just want to add that I consider SAS to be much superior in terms of programming flexibility. This would include data manipulation, SQL-types of queries, merges, joins, transposing data sets, using arrays and loops, macro facilities in which entire procedures can be invoked iteratively, creation of various types of output files in many formats (using the ODS facility), using functions, being able to structure your code in any way that seems appropriate (rather than being confined to point-and-click format or other rigid formats), and so on—basically programming tasks rather than statistical ones."

I don't know what version of SPSS Chuck refers to but he does everyone disservice by being terribly misinformed and passing on this disinformation with respect to the capabilities of SPSS. I surmise pre version 4 (1990ish) or SPSSPC+ ???
SPSS has supported ALL OF THE ABOVE for as long as I can remember!!!
Please be informed before opening your pie hole?
SQL-types of queries ->
merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
using arrays and loops-> VECTOR, LOOP, DO REPEAT
macro facility-> DEFINE !ENDDEFINE
creation of various types of output files in many formats -> OMS, OUTPUT SAVE
Code can be structured in any way one sees fit.
In NO WAY re people confined to "Point and Click".
Yeah, Basically PROGRAMMING!!!!!!!
In fact, I challenge Chuck to come up with any task doable in SAS that can't be done (by someone like me) more elegantly using SPSS!!

Andy W

Re: FW: SPSS Statistics Survey

Totally agree with Bruce,

Re the prices, if you go to this page, http://www-01.ibm.com/software/analytics/spss/products/statistics/, click on a particular option, then there is a big blue button on the right to view prices.

Here is the page for US prices for the just the base package. For a perpetual, single use license the quote is currently $2,610.

https://www-112.ibm.com/software/howtobuy/buyingtools/paexpress/Express?P0=E1&part_number=D0EJ9LL,D0EEILL,D0ELQLL,D0EEFLL&catalogLocale=en_US&Locale=en_US&country=USA&PT=jsp&CC=USA&VP=&TACTICS=&S_TACT=&S_CMP=&brand=SSUG68

(Pro-tip - if you need a quote calling works much faster than using the online forms.)

Stata is cheaper by around $1,000, see http://www.stata.com/order/new/bus/single-user-licenses/dl/. Also note that Stata has no additional add-ons you need to buy for statistical procedures. (The statistics standard license for SPSS will run you near $6,000 currently.)

(The only things not in SPSS base that I use ever are GENLIN, MIXED and GENLINMIXED, and these I use so little I haven't recommended getting them at when it has come up at a few of my past jobs.)

One thing that ArcGIS does now that is nice is that they have a program for at home, personal (non-profit) use. It is just $100 a year, and includes the most usual statistical/data management add-ons academics use for their work. (And ESRI's price scheme is more comparable to SPSS than to Stata.) You have to resign up and redownload the newer version software every year, but I would totally sign up for that if SPSS had something comparable. [Students can come close to doing this currently by buying yearly student licenses from various vendors.]

(SAS's academic version seems to be entirely in the cloud from what I can tell - and I imagine SAS limits the type of procedures that can be done as well.)

Andy W
apwheele@gmail.com
http://andrewpwheeler.wordpress.com/

Ryan

Re: SPSS Statistics Survey

In reply to this post by David Marso

I would suggest no one take on David's challenge. :-)

The only criticism I have, and have always had is that SPSS is lagging behind with its statistical capabilities compared to SAS. Having said that, with every release and with R extensions, SPSS is catching up.

Briefly, I'd like to see the TEST sub-command become available in GENLIN and GENLINMIXED. I would also like to see something similar to the NLMIXED procedure in SAS. In MIXED, it would also be helpful to choose to different degrees of freedom options. Finally, I'd be thrilled to see other estimation methods and techniques offered in various procedures, including Bayesian estimation (eg MCMC) among others.

Ryan

Sent from my iPhone

> On Jan 30, 2015, at 2:58 AM, David Marso <[hidden email]> wrote:
>
> Chuck wrote:
> "I just want to add that I consider SAS to be much superior in terms of
> programming flexibility. This would include data manipulation, SQL-types of
> queries, merges, joins, transposing data sets, using arrays and loops, macro
> facilities in which entire procedures can be invoked iteratively, creation
> of various types of output files in many formats (using the ODS facility),
> using functions, being able to structure your code in any way that seems
> appropriate (rather than being confined to point-and-click format or other
> rigid formats), and so on—basically programming tasks rather than
> statistical ones."
>
> I don't know what version of SPSS Chuck refers to but he does everyone
> disservice by being terribly misinformed and passing on this disinformation
> with respect to the capabilities of SPSS. I surmise pre version 4 (1990ish)
> or SPSSPC+ ???
> SPSS has supported *ALL OF THE ABOVE *for as long as I can remember!!!
> *Please be informed before opening your pie hole?*
> SQL-types of queries ->
> merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
> transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
> using arrays and loops-> VECTOR, LOOP, DO REPEAT
> macro facility-> DEFINE !ENDDEFINE
> creation of various types of output files in many formats -> OMS, OUTPUT
> SAVE
> Code can be structured in any way one sees fit.
> In NO WAY re people confined to "Point and Click".
> *Yeah, Basically PROGRAMMING!!!!!!!*
> In fact, *I challenge Chuck to come up with any task doable in SAS that
> can't be done (by someone like me) more elegantly using SPSS*!!
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SPSS-Statistics-Survey-tp5728513p5728532.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

Bruce Weaver

Re: FW: SPSS Statistics Survey

Administrator

In reply to this post by David Marso

David, perhaps you should post your challenge to comp.soft-sys.sas. ;-)

https://groups.google.com/forum/#!forum/comp.soft-sys.sas

I suppose we'd have to convene a panel of impartial judges from the R and Stata lists.

David Marso wrote

Chuck wrote:
"I just want to add that I consider SAS to be much superior in terms of programming flexibility. This would include data manipulation, SQL-types of queries, merges, joins, transposing data sets, using arrays and loops, macro facilities in which entire procedures can be invoked iteratively, creation of various types of output files in many formats (using the ODS facility), using functions, being able to structure your code in any way that seems appropriate (rather than being confined to point-and-click format or other rigid formats), and so on—basically programming tasks rather than statistical ones."

I don't know what version of SPSS Chuck refers to but he does everyone disservice by being terribly misinformed and passing on this disinformation with respect to the capabilities of SPSS. I surmise pre version 4 (1990ish) or SPSSPC+ ???
SPSS has supported ALL OF THE ABOVE for as long as I can remember!!!
Please be informed before opening your pie hole?
SQL-types of queries ->
merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
using arrays and loops-> VECTOR, LOOP, DO REPEAT
macro facility-> DEFINE !ENDDEFINE
creation of various types of output files in many formats -> OMS, OUTPUT SAVE
Code can be structured in any way one sees fit.
In NO WAY re people confined to "Point and Click".
Yeah, Basically PROGRAMMING!!!!!!!
In fact, I challenge Chuck to come up with any task doable in SAS that can't be done (by someone like me) more elegantly using SPSS!!

Ryan

Re: SPSS Statistics Survey

> On Jan 30, 2015, at 9:20 PM, Bruce Weaver <[hidden email]> wrote:
>
> David, perhaps you should post your challenge to comp.soft-sys.sas. ;-)
>
> https://groups.google.com/forum/#!forum/comp.soft-sys.sas
>
> I suppose we'd have to convene a panel of impartial judges from the R and
> Stata lists.
>
>
>
> David Marso wrote
>> Chuck wrote:
>> "I just want to add that I consider SAS to be much superior in terms of
>> programming flexibility. This would include data manipulation, SQL-types
>> of queries, merges, joins, transposing data sets, using arrays and loops,
>> macro facilities in which entire procedures can be invoked iteratively,
>> creation of various types of output files in many formats (using the ODS
>> facility), using functions, being able to structure your code in any way
>> that seems appropriate (rather than being confined to point-and-click
>> format or other rigid formats), and so on—basically programming tasks
>> rather than statistical ones."
>>
>> I don't know what version of SPSS Chuck refers to but he does everyone
>> disservice by being terribly misinformed and passing on this
>> disinformation with respect to the capabilities of SPSS. I surmise pre
>> version 4 (1990ish) or SPSSPC+ ???
>> SPSS has supported
> *
>> ALL OF THE ABOVE
> *
>> for as long as I can remember!!!
> *
>> Please be informed before opening your pie hole?
> *
>> SQL-types of queries ->
>> merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
>> transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
>> using arrays and loops-> VECTOR, LOOP, DO REPEAT
>> macro facility-> DEFINE !ENDDEFINE
>> creation of various types of output files in many formats -> OMS, OUTPUT
>> SAVE
>> Code can be structured in any way one sees fit.
>> In NO WAY re people confined to "Point and Click".
> *
>> Yeah, Basically PROGRAMMING!!!!!!!
> *
>> In fact,
> *
>> I challenge Chuck to come up with any task doable in SAS that can't be
>> done (by someone like me) more elegantly using SPSS
> *
>> !!
>
>
>
>
>
> -----
> --
> Bruce Weaver
> [hidden email]
> http://sites.google.com/a/lakeheadu.ca/bweaver/
>
> "When all else fails, RTFM."
>
> NOTE: My Hotmail account is not monitored regularly.
> To send me an e-mail, please use the address shown above.
>
> --
> View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SPSS-Statistics-Survey-tp5728513p5728537.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

Jon K Peck

Re: SPSS Statistics Survey

Analyze > Simulation > Create Simulated Data
Then you have several choices including simulating from a pmml model, simulating from equations you enter, and create simulated data without a model. You can specify the individual variable distributions and a correlation matrix or use one fit to a dataset.

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: 720-342-5621

From: [hidden email]
To: [hidden email]
Date: 01/30/2015 09:25 PM
Subject: Re: [SPSSX-L] SPSS Statistics Survey
Sent by: "SPSSX(r) Discussion" <[hidden email]>

One capability I find quite useful in SAS is generating data which arise from a MVN distribution with a specified covariance structure to conform to all types of parameterizations of statistical models within the matrix language environment offered in SAS/IML. I seem to recall that this type of data could not be generated IN the matrix command in SPSS, but I could be wrong. Then again, there is a simulation procedure offered in SPSS now. Ryan Sent from my iPhone > On Jan 30, 2015, at 9:20 PM, Bruce Weaver <[hidden email]> wrote: > > David, perhaps you should post your challenge to comp.soft-sys.sas. ;-) > > https://groups.google.com/forum/#!forum/comp.soft-sys.sas> > I suppose we'd have to convene a panel of impartial judges from the R and > Stata lists. > > > > David Marso wrote >> Chuck wrote: >> "I just want to add that I consider SAS to be much superior in terms of >> programming flexibility. This would include data manipulation, SQL-types >> of queries, merges, joins, transposing data sets, using arrays and loops, >> macro facilities in which entire procedures can be invoked iteratively, >> creation of various types of output files in many formats (using the ODS >> facility), using functions, being able to structure your code in any way >> that seems appropriate (rather than being confined to point-and-click >> format or other rigid formats), and so on—basically programming tasks >> rather than statistical ones." >> >> I don't know what version of SPSS Chuck refers to but he does everyone >> disservice by being terribly misinformed and passing on this >> disinformation with respect to the capabilities of SPSS. I surmise pre >> version 4 (1990ish) or SPSSPC+ ??? >> SPSS has supported > * >> ALL OF THE ABOVE > * >> for as long as I can remember!!! > * >> Please be informed before opening your pie hole? > * >> SQL-types of queries -> >> merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support. >> transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS >> using arrays and loops-> VECTOR, LOOP, DO REPEAT >> macro facility-> DEFINE !ENDDEFINE >> creation of various types of output files in many formats -> OMS, OUTPUT >> SAVE >> Code can be structured in any way one sees fit. >> In NO WAY re people confined to "Point and Click". > * >> Yeah, Basically PROGRAMMING!!!!!!! > * >> In fact, > * >> I challenge Chuck to come up with any task doable in SAS that can't be >> done (by someone like me) more elegantly using SPSS > * >> !! > > > > > > ----- > -- > Bruce Weaver > [hidden email] >http://sites.google.com/a/lakeheadu.ca/bweaver/> > "When all else fails, RTFM." > > NOTE: My Hotmail account is not monitored regularly. > To send me an e-mail, please use the address shown above. > > -- > View this message in context:http://spssx-discussion.1045642.n5.nabble.com/SPSS-Statistics-Survey-tp5728513p5728537.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
===================== 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

David Marso

Re: FW: SPSS Statistics Survey

Administrator

In reply to this post by Bruce Weaver

I'd feel more comfortable having you or some other 'neutral' toss it over the wall and see who wants to come out and play. I wouldn't want to come off too cocky jumping into the lion's den ;-)
------

Bruce Weaver wrote

David, perhaps you should post your challenge to comp.soft-sys.sas. ;-)

https://groups.google.com/forum/#!forum/comp.soft-sys.sas

I suppose we'd have to convene a panel of impartial judges from the R and Stata lists.

David Marso wrote

Chuck wrote:
"I just want to add that I consider SAS to be much superior in terms of programming flexibility. This would include data manipulation, SQL-types of queries, merges, joins, transposing data sets, using arrays and loops, macro facilities in which entire procedures can be invoked iteratively, creation of various types of output files in many formats (using the ODS facility), using functions, being able to structure your code in any way that seems appropriate (rather than being confined to point-and-click format or other rigid formats), and so on—basically programming tasks rather than statistical ones."

I don't know what version of SPSS Chuck refers to but he does everyone disservice by being terribly misinformed and passing on this disinformation with respect to the capabilities of SPSS. I surmise pre version 4 (1990ish) or SPSSPC+ ???
SPSS has supported ALL OF THE ABOVE for as long as I can remember!!!
Please be informed before opening your pie hole?
SQL-types of queries ->
merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.
transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS
using arrays and loops-> VECTOR, LOOP, DO REPEAT
macro facility-> DEFINE !ENDDEFINE
creation of various types of output files in many formats -> OMS, OUTPUT SAVE
Code can be structured in any way one sees fit.
In NO WAY re people confined to "Point and Click".
Yeah, Basically PROGRAMMING!!!!!!!
In fact, I challenge Chuck to come up with any task doable in SAS that can't be done (by someone like me) more elegantly using SPSS!!

Richard Ristow

Re: FW: SPSS Statistics Survey

In reply to this post by David Marso

At 02:58 AM 1/30/2015, David Marso quoted 'Chuck', and responded; I
interleave my own thoughts:

Chuck:
>>I consider SAS to be much superior in terms of programming flexibility.
>>This would include ... SQL-types of queries, merges, joins,
David:
>SQL-types of queries ->
>merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL support.

SPSS's capabilities for merging and joining have been much, much
nearer to SAS's since MATCH FILES and ADD FILES were added in release
SPSS-X in, I believe, 1983. Nevertheless, SAS's SET and MERGE
statements have features that I miss badly in SPSS's GET FILE/ADD
FILE and MATCH FILES statements, including:

. Separate KEEP and DROP lists for each input. Those make selecting
sets of variables from the inputs much neater (variables with
conflicting names from different inputs can simply be dropped if you
don't need them, rather than renamed and then dropped) and much more
readable (the SET or MERGE statement makes clear which variables come
from which inputs, especially if one uses KEEP lists.)

. Option END=, which names a variable that flags the last case being
built, and can be used to trigger end-of-input processing code.

. FIRST and LAST can be triggered separately for each variable in a
multi-variable BY key, so you can have different code for different
break levels.

. And SAS (unless my recollection fails me badly) has *nothing* like
SPSS's rule that makes different-length strings type-incompatible for
merging -- I find that maddening particularly because it's so
unnecessary, and don't feel mollified by the existence of Python-y
work-arounds.

Then, there are SQL-type queries. I never wrote a lot of SQL, but
have found SAS's PROC SQL invaluable for a couple of things:
many-to-many joins of a file with itself, and transitive-closure
logic. More on those in a separate note, I hope.

I haven't used STAR JOIN except briefly, in a beta test, but I found
it a major disappointment. Its sort-of-but-not-really SQL syntax
seems like a half-considered "Me, too!" to PROC SQL; and a lot of
what it does offer -- merging unsorted inputs, *not* having that
maddening string-length incompatibility -- would do as well or better
as upgrades to MATCH FILES/ADD FILES.

Chuck:
>>transposing data sets,
David:
>transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS

I always found FLIP limited to the point of uselessness; it was a
great relief to get to SAS and its PROC TRANSPOSE. But VARSTOCASES
and CASESTOVARS are very good, and I doubt that PROC TRANSPOSE has
much if any advantage over them. (I haven't used PROC TRANSPOSE for
quite a while, so I can't make detailed comparisons.)

However, SAS has the OUTPUT statement that lets you programmatically
determine when a case is completed -- closely analogous to END CASE
in INPUT PROGRAM. You can use OUTPUT logic to do quite complicated
transpose-like programming that would be very difficult in SPSS.

(By the by, INPUT PROGRAM is another feature that maddens me. It's
very useful, especially with wise use of END CASE; but then, you
can't use any of that logic with data that's already in SPSS. How
about allowing GET FILE/ADD FILES/MATCH FILE in something analogous
to an INPUT PROGRAM, with END CASE or something like it?)

Chuck:
>>using arrays and loops,
David
>using arrays and loops-> VECTOR, LOOP, DO REPEAT

Exactly so. Though SAS arrays don't have to have their members
contiguous in the file, and non-contiguous arrays are very handy sometimes.

Chuck:
>>macro facilities in which entire procedures can be invoked iteratively,
David:
>macro facility-> DEFINE !ENDDEFINE

Again, exactly so. Though SAS's macro facility is also, in part, a
scripting facility; that is, it can interact with a running program.
That gives it a power, flexibility, and complexity that make it
notoriously complicated to use; but its ability to, for example,
terminate a loop on the basis of some count from a procedure are
useful for effects like (here we go again) transitive closure.

SPSS's real answer to SAS's macro-scripting facility is the Python
interface, which I haven't used enough to write about. I think there
are very few script-like applications in SAS that couldn't be done as
well in SPSS-Python.

Other points:

. SAS's PROC SUMMARY is more powerful than AGGREGATE, but a lot
clumsier. I find AGGREGATE much neater and clearer to use.

. SAS's PROC FORMATS gives most of the capabilities of SPSS's VALUE
LABELS and RECODE, at the expense of clumsier coding: the format
definitions (which can be made to function as RECODE definitions)
have to be written and saved separately from the main program, and
stored separately from the data.

. SPSS datasets now give much of the flexibility for creating and
using temporary files, that SAS has always had in its _WORK_ data
sets. I find SPSS's deliberately making datasets and .SAV files
code-incompatible to be maddening sometimes: you can't COPY to a .SAV
file and you can't SAVE or XSAVE to a dataset, and it can be very
hard to write code that will work with both .SAV files and datasets.

. And, of course, there's pricing policy. SAS is licensed by the
year, at a very high rate. Use SAS and, any year you don't renew your
license, all your code and (unless you're careful) much of your data
is immediately useless. I'd love to use SAS more, but it essentially
can't be done unless you're affiliated with an institution that has a
long-term commitment to maintaining a SAS site license.

=====================
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

Garry Gelade

Re: SPSS Statistics Survey

In reply to this post by Ryan

Yes, Bayesian capabilities would be brilliant

Garry

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of [hidden email]
Sent: 30 January 2015 14:21
To: [hidden email]
Subject: Re: SPSS Statistics Survey

I would suggest no one take on David's challenge. :-)

The only criticism I have, and have always had is that SPSS is lagging behind with its statistical capabilities compared to SAS. Having said that, with every release and with R extensions, SPSS is catching up.

Briefly, I'd like to see the TEST sub-command become available in GENLIN and GENLINMIXED. I would also like to see something similar to the NLMIXED procedure in SAS. In MIXED, it would also be helpful to choose to different degrees of freedom options. Finally, I'd be thrilled to see other estimation methods and techniques offered in various procedures, including Bayesian estimation (eg MCMC) among others.

Ryan

Sent from my iPhone

> On Jan 30, 2015, at 2:58 AM, David Marso <[hidden email]> wrote:
>
> Chuck wrote:
> "I just want to add that I consider SAS to be much superior in terms
> of programming flexibility. This would include data manipulation,
> SQL-types of queries, merges, joins, transposing data sets, using
> arrays and loops, macro facilities in which entire procedures can be
> invoked iteratively, creation of various types of output files in many
> formats (using the ODS facility), using functions, being able to
> structure your code in any way that seems appropriate (rather than
> being confined to point-and-click format or other rigid formats), and
> so on—basically programming tasks rather than statistical ones."
>
> I don't know what version of SPSS Chuck refers to but he does everyone
> disservice by being terribly misinformed and passing on this
> disinformation with respect to the capabilities of SPSS. I surmise
> pre version 4 (1990ish) or SPSSPC+ ???
> SPSS has supported *ALL OF THE ABOVE *for as long as I can remember!!!
> *Please be informed before opening your pie hole?* SQL-types of
> queries -> merges/joins -> MATCH FILES, ADD FILES, STAR JOIN + SQL
> support.
> transposing data sets -> FLIP, VARSTOCASES, CASESTOVARS using arrays
> and loops-> VECTOR, LOOP, DO REPEAT macro facility-> DEFINE !ENDDEFINE
> creation of various types of output files in many formats -> OMS,
> OUTPUT SAVE Code can be structured in any way one sees fit.
> In NO WAY re people confined to "Point and Click".
> *Yeah, Basically PROGRAMMING!!!!!!!*
> In fact, *I challenge Chuck to come up with any task doable in SAS
> that can't be done (by someone like me) more elegantly using SPSS*!!
>
>
>
> -----
> Please reply to the list and not to my personal email.
> Those desiring my consulting or training services please feel free to email me.
> ---
> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
> --
> View this message in context:
> http://spssx-discussion.1045642.n5.nabble.com/SPSS-Statistics-Survey-t
> p5728513p5728532.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

=====================
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

Richard Ristow

Re: FW: SPSS Statistics Survey

In reply to this post by David Marso

At 02:58 AM 1/30/2015, David Marso wrote:

>In fact, *I challenge Chuck to come up with any task doable in SAS
>that can't be done (by someone like me) more elegantly using SPSS*!!

I've done a couple of big projects in SAS that I'd have found
difficult to impossible in SPSS; more on those, later. Here's a
simple snippet that I've used frequently in SAS, that I think goes
better there than it does in SPSS. It's a template for checking a
file for errors and inconsistencies, and printing a report; see
Notes, following the code. (And, apologies for any SAS syntax errors.
I'm rusty in SAS, and don't have a copy to check the code with.)

DATA _NULL_;
SET To_Check END=Done;

RETAIN TotCases ErrCases TotErrors 0;

FILE PRINT LinesLeft = SPACE;

TotCases = TotCases + 1;
ErrFound = 0;

/* Error checks; put as many as desired */

IF Alpha < 1 ^ Alpha > 17
THEN DO;
LINK HdrPrint;
PUT ' Alpha not between 1 and 17: ' Alpha=;
END;

IF Beta > Gamma
THEN DO;
LINK HdrPrint;
PUT ' Beta is larger than Gamma: ' Beta= Gamma=;
END;

/* Summary statistics, at end */

IF Done
THEN DO;
Put 'Checked ' TotCases ' cases';
PUT 'Found ' TotErrors ' errors, in ' ErrCases ' cases';
END;

RETURN;
HdrPrint:
IF ~ErrFound
THEN DO;
ErrFound = 1;
ErrCases = ErrCases + 1;
IF SPACE < 6 THEN PUT _PAGE_;
ELSE PUT ' ';
PUT 'Case ' IDvar1 Idvar2;
END;
TotErrors = TotErrors + 1;
RETURN;
/* --------------------------------------------------------- */

NOTES:
This uses the following SAS features that don't have SPSS equivalents:

+ LINK/RETURN to call HdrPrint as a subroutine, so its code doesn't
have to be repeated as part of every tests. (In SPSS, however, the
HdrPrint code could be wrapped in a macro, so it's easy to insert it
in each test:

DO IF NOT RANGE(Alpha,1,17).
!HdrPrint
PRINT ' Alpha not in range 1-17: ' Alpha
END IF.

+ END=Done option on SET, to trigger end-of-run report printing

+ Better control of printer formatting, with LINESLEFT=Space and PUT
_PAGE_ to keep case headers together with their corresponding error messages.

=====================
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

Jon K Peck

Re: FW: SPSS Statistics Survey

I must point out that this data validation would have been much easier in Statistics if you used the VALIDATEDATA procedure and its nice gui to define all the validation rules and run the tests. And the output automatically comes out as easily readable tables.

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: 720-342-5621

From: Richard Ristow <[hidden email]>
To: [hidden email]
Date: 02/02/2015 10:28 AM
Subject: Re: [SPSSX-L] FW: SPSS Statistics Survey
Sent by: "SPSSX(r) Discussion" <[hidden email]>

At 02:58 AM 1/30/2015, David Marso wrote: >In fact, *I challenge Chuck to come up with any task doable in SAS >that can't be done (by someone like me) more elegantly using SPSS*!! I've done a couple of big projects in SAS that I'd have found difficult to impossible in SPSS; more on those, later. Here's a simple snippet that I've used frequently in SAS, that I think goes better there than it does in SPSS. It's a template for checking a file for errors and inconsistencies, and printing a report; see Notes, following the code. (And, apologies for any SAS syntax errors. I'm rusty in SAS, and don't have a copy to check the code with.) DATA _NULL_; SET To_Check END=Done; RETAIN TotCases ErrCases TotErrors 0; FILE PRINT LinesLeft = SPACE; TotCases = TotCases + 1; ErrFound = 0; /* Error checks; put as many as desired */ IF Alpha < 1 ^ Alpha > 17 THEN DO; LINK HdrPrint; PUT ' Alpha not between 1 and 17: ' Alpha=; END; IF Beta > Gamma THEN DO; LINK HdrPrint; PUT ' Beta is larger than Gamma: ' Beta= Gamma=; END; /* Summary statistics, at end */ IF Done THEN DO; Put 'Checked ' TotCases ' cases'; PUT 'Found ' TotErrors ' errors, in ' ErrCases ' cases'; END; RETURN; HdrPrint: IF ~ErrFound THEN DO; ErrFound = 1; ErrCases = ErrCases + 1; IF SPACE < 6 THEN PUT _PAGE_; ELSE PUT ' '; PUT 'Case ' IDvar1 Idvar2; END; TotErrors = TotErrors + 1; RETURN; /* --------------------------------------------------------- */ NOTES: This uses the following SAS features that don't have SPSS equivalents: + LINK/RETURN to call HdrPrint as a subroutine, so its code doesn't have to be repeated as part of every tests. (In SPSS, however, the HdrPrint code could be wrapped in a macro, so it's easy to insert it in each test: DO IF NOT RANGE(Alpha,1,17). !HdrPrint PRINT ' Alpha not in range 1-17: ' Alpha END IF. + END=Done option on SET, to trigger end-of-run report printing + Better control of printer formatting, with LINESLEFT=Space and PUT _PAGE_ to keep case headers together with their corresponding error messages. ===================== 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