Software review time

[Maguin, Eugene]

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

=====================
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]

Gene is a regular and valuable member of this list. He is likely to be able to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school to learn the other software multiplied by the hourly cost of their compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

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

[Mark Webb-5]

Gene
They have a 30 day free download offer on their site. Try it out for yourself.
When I thought SPSS was pushing the price a bit [in South Africa] I converted to a perpetual license at a once-off cost.
I miss out on new developments though - to solve this problem I learnt how to use R.
Normally anything new in SPSS will be covered somewhere in R.
So now both are free to me & I'm finding this combination very useful.
 
On 2011/04/14 11:47 PM, Art Kendall wrote:

Gene is a regular and valuable member of this list. He is likely to be able to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school to learn the other software multiplied by the hourly cost of their compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

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

[Thomas MacFarland]

Everyone:

 

For those who are just starting (whether as a student or at an office), where there is no existing overhead to consider, then it may not make too much difference if the selection is Product X or Product Y.  Most major statistics packages are fine.

 

However, for anyone who has existing routines, the cost of rewriting everything must be considered, which is far more expensive that the mere cost of learning a new package.

 

In our office I can not even begin to imagine what it would cost in man hours to switch (That is to say, organize data to meet the new package requirements, write new syntax, test outcomes, etc.), from SPSS to anything else.  And remember, many analyses are retrospective so it would be necessary to work with old data as well as anything new.

 

And, for SPSS specifically – I sure would not want to see the Custom Tables feature go away.

 

Comment:  For many new routines, where there is no established history in our office, I’ve been using R lately.  It certainly deserves attention from anyone who has good skills with syntax and programming in general.

 

Best wishes.

 

Tom

 

-----

Thomas W. MacFarland, Ed.D.
Senior Research Associate; Institutional Effectiveness and Associate Professor
Nova Southeastern University
Voice 954-262-5395  Fax [phone]  [hidden email]

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
Sent: Thursday, April 14, 2011 5:47 PM
To: [hidden email]
Subject: Re: Software review time

 

Gene is a regular and valuable member of this list. He is likely to be able to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school to learn the other software multiplied by the hourly cost of their compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per

year (price probably applies to just this year). Systat is being mentioned

as a replacement. I'd like to hear off-list or on-list about any published

comparisons of spss and sysstat. As well as the experiences of people who

have both. I've looked at sysstat's website and have the pdf of version 13

features, so I'm not so much interested in statistical tests per se. I

figure that probably 70-80% of my spss time is data management chores. And,

I write syntax, 'never' drop-down menus. So, I'd particulary like to hear

comments pertaining to the editors, the syntax language and it breadth,

functionality, and flexibility and its inclusion of somewhat recently

introduced spss commands. How is the output file structured: Ascii/unicode

text or something more spss-like. I assume computational accuracy is not an

issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have

a lot of spss experience. I'm mainly concerned with how easily my knowledge

and experience will slide over if spss is abandoned.

Gene Maguin

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

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

[J P-6]

SPSS has been an enornous part of my professional life, having used it almost 20 years now. That said, I see more and more folks switching to Stata. I have switched to Stata for my personal conusulting work because 1) it makes use of menu & syntax commands 2) is feature rich 'out-of-the-box' (no $$ for extra modules), and 3) I can purchase a perpetual license at an affordable  price, and 4) the software is not upgraded every foggy morning. Plus it it seems to have the flexability of R without  the steep learning curve. 

 

Best,

John

 

 

---

From: Mark Webb <[hidden email]>
To: [hidden email]
Sent: Thu, April 14, 2011 11:41:06 PM
Subject: Re: Software review time

Gene
They have a 30 day free download offer on their site. Try it out for yourself.
When I thought SPSS was pushing the price a bit [in South Africa] I converted to a perpetual license at a once-off cost.
I miss out on new developments though - to solve this problem I learnt how to use R.
Normally anything new in SPSS will be covered somewhere in R.
So now both are free to me & I'm finding this combination very useful.
 
On 2011/04/14 11:47 PM, Art Kendall wrote:

Gene is a regular and valuable member of this list. He is likely to be able to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school to learn the other software multiplied by the hourly cost of their compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

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

[Bruce Weaver]

I agree with most of what Tom says.  Where I differ a bit is on his comment that most major packages are fine/comparable.  I *think* that if I was a youngster starting from scratch, I would be quite tempted to go with Stata over SPSS.  Some of the reasons are summarized in a nice technical report available on the UCLA Computing website.

    http://www.ats.ucla.edu/stat/technicalreports/

One thing I like about this report is that the author has pretty extensive experience with all 3 of the packages being discussed.  That distinguishes it from the more common situation where authors know a lot about one of the packages, but not very much about the others--so it degenerates into "my favorite package is better than yours".  ;-)

Cheers,
Bruce

p.s. - My half-baked plan for after retirement (still several years off) is to learn R.  At that point, I'll not be able to get SPSS via the university site license, and I don't fancy paying the full whack for it!

Dr. Thomas W. MacFarland wrote

Everyone:



For those who are just starting (whether as a student or at an office),
where there is no existing overhead to consider, then it may not make too
much difference if the selection is Product X or Product Y.  Most major
statistics packages are fine.



However, for anyone who has existing routines, the cost of rewriting
everything must be considered, which is far more expensive that the mere
cost of learning a new package.



In our office I can not even begin to imagine what it would cost in man
hours to switch (That is to say, organize data to meet the new package
requirements, write new syntax, test outcomes, etc.), from SPSS to anything
else.  And remember, many analyses are retrospective so it would be
necessary to work with old data as well as anything new.



And, for SPSS specifically - I sure would not want to see the Custom Tables
feature go away.



Comment:  For many new routines, where there is no established history in
our office, I've been using R lately.  It certainly deserves attention from
anyone who has good skills with syntax and programming in general.



Best wishes.



Tom



-----

Thomas W. MacFarland, Ed.D.
Senior Research Associate; Institutional Effectiveness and Associate
Professor
Nova Southeastern University
Voice 954-262-5395  Fax [phone]  [hidden email]



  _____

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art
Kendall
Sent: Thursday, April 14, 2011 5:47 PM
To: [hidden email]
Subject: Re: Software review time



Gene is a regular and valuable member of this list. He is likely to be able
to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school
to learn the other software multiplied by the hourly cost of their
compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

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

[Bruce Weaver]

After posting that last message, I re-read some parts of Michael Mitchell's report.  The section on data management is incomplete, but I was struck by the second paragraph.

--- excerpt from data management section ---

I will, however, say that I believe that the packages can differ substantially in their strengths with respect
to data management tasks. Even if you have made a selection of a package that is most suitable for your
data analysis, I would encourage you to separately assess the strengths and weaknesses of the packages for your data management and if a different package is most useful for your data management, then use one package for your data management and use a different package for your data analysis. Converting data from one package to another is so easy these days, it makes good sense to select the best tool for data management and the best tool for your data analysis, even if the tools are different. The cost of converting your data from one format to another is so small compared to the benefit that you can get from selecting the best tool at each stage of your analysis.

--- end of excerpt ---

So one option for people who are well-versed in SPSS might be to get a permanent license for the base module, and use it for data management, while using something else (e.g., R or Stata) for analyses that are not included with the base module.  Given how much time is spent on data management relative to analysis, this might be a pretty nice way to have the best of both worlds.

Has anyone purchased the base module lately?  How much is it going for?

Bruce Weaver wrote

I agree with most of what Tom says.  Where I differ a bit is on his comment that most major packages are fine/comparable.  I *think* that if I was a youngster starting from scratch, I would be quite tempted to go with Stata over SPSS.  Some of the reasons are summarized in a nice technical report available on the UCLA Computing website.

    http://www.ats.ucla.edu/stat/technicalreports/

One thing I like about this report is that the author has pretty extensive experience with all 3 of the packages being discussed.  That distinguishes it from the more common situation where authors know a lot about one of the packages, but not very much about the others--so it degenerates into "my favorite package is better than yours".  ;-)

Cheers,
Bruce

p.s. - My half-baked plan for after retirement (still several years off) is to learn R.  At that point, I'll not be able to get SPSS via the university site license, and I don't fancy paying the full whack for it!

Dr. Thomas W. MacFarland wrote

Everyone:



For those who are just starting (whether as a student or at an office),
where there is no existing overhead to consider, then it may not make too
much difference if the selection is Product X or Product Y.  Most major
statistics packages are fine.



However, for anyone who has existing routines, the cost of rewriting
everything must be considered, which is far more expensive that the mere
cost of learning a new package.



In our office I can not even begin to imagine what it would cost in man
hours to switch (That is to say, organize data to meet the new package
requirements, write new syntax, test outcomes, etc.), from SPSS to anything
else.  And remember, many analyses are retrospective so it would be
necessary to work with old data as well as anything new.



And, for SPSS specifically - I sure would not want to see the Custom Tables
feature go away.



Comment:  For many new routines, where there is no established history in
our office, I've been using R lately.  It certainly deserves attention from
anyone who has good skills with syntax and programming in general.



Best wishes.



Tom



-----

Thomas W. MacFarland, Ed.D.
Senior Research Associate; Institutional Effectiveness and Associate
Professor
Nova Southeastern University
Voice 954-262-5395  Fax [phone]  [hidden email]



  _____

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art
Kendall
Sent: Thursday, April 14, 2011 5:47 PM
To: [hidden email]
Subject: Re: Software review time



Gene is a regular and valuable member of this list. He is likely to be able
to learn other software more quickly than most.

I wonder how much additional time it would take the users at Gene's school
to learn the other software multiplied by the hourly cost of their
compensation packages?
each
totalcost = (hours * compensation) + purchaseprice.

Art

On 4/14/2011 5:09 PM, Gene Maguin wrote:

We just found out that spss is going to be charging $300 per license per
year (price probably applies to just this year). Systat is being mentioned
as a replacement. I'd like to hear off-list or on-list about any published
comparisons of spss and sysstat. As well as the experiences of people who
have both. I've looked at sysstat's website and have the pdf of version 13
features, so I'm not so much interested in statistical tests per se. I
figure that probably 70-80% of my spss time is data management chores. And,
I write syntax, 'never' drop-down menus. So, I'd particulary like to hear
comments pertaining to the editors, the syntax language and it breadth,
functionality, and flexibility and its inclusion of somewhat recently
introduced spss commands. How is the output file structured: Ascii/unicode
text or something more spss-like. I assume computational accuracy is not an
issue.

Please, no spss bashing. Probably, like a lot of people on this list, I have
a lot of spss experience. I'm mainly concerned with how easily my knowledge
and experience will slide over if spss is abandoned.

Gene Maguin

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

[Jon K Peck]

It should be noted that the UCLA article is very old.  Regarding SPSS, it refers to version 14 - five full releases ago, and it therefore does not reflect the many things that have changed since in the software - in statistical capabilities, graphics, data management, programmability, user interface, the Community, or business terms.  In fact, it doesn't even reflect important things that happened in V14.  I don't want to start a debate, but the article is nowhere near current.

Jon Peck
Senior Software Engineer, IBM
[hidden email]
312-651-3435



From:        Bruce Weaver <[hidden email]>
To:        [hidden email]
Date:        04/15/2011 09:33 AM
Subject:        Re: [SPSSX-L] Software review time
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

After posting that last message, I re-read some parts of Michael Mitchell's
report.  The section on data management is incomplete, but I was struck by
the second paragraph.