inexpensive 'home' version?

[Steve Simon, P.Mean Consulting]

I just noticed all the discussion about how you NEVER should compute the
mean of an ordinal variable. While I can't disagree with many of the
fine alternatives being suggested, I would mention that it is dangerous
to say "never".

Most people (including many statisticians) use a statistic called grade
point average (GPA) in their hiring decisions. This is clearly an
average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
to pretend that these numbers can be averaged? Well it implies that two
C's are equivalent to an A and an F. Still, the GPA doesn't do too much
violence to the data, and I think all of us would agree that a 3.5 GPA
is better than a 2.5 GPA.

Whether you should average or not depends a lot on the distribution of
the ordinal categories, the number of levels of the ordinal variable,
what the levels of the ordinal variable represent, and what the goals of
the statistical summary are.

I'd argue that the average of a 6 point variable is a simple summary
that is useful for descriptive statistics or informal comparisons. I'd
save the more complex alternatives for larger projects where a "quick
and dirty" approximation would be inappropriate.

I talk a bit more about this and point out the problems with a commonly
used alternative for ordinal data, rank-based methods, at my old website:
  * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
--
Steve Simon, Standard Disclaimer
Sign up for The Monthly Mean, the newsletter that
dares to call itself "average" at www.pmean.com/news
"Data entry and data management issues with examples
in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
Free webinar. Details at www.pmean.com/webinars

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

Nice note on your website, Steve.  I have two comments.

1. I believe that Likert would only have used the term "scale" after summing or averaging over several items.  I.e., it takes more than one Likert-type item to make a scale.

2. If I'm not mistaken, Likert tried several more complex methods of assigning numbers to the ordinal categories (like your -3 -1 0 1 3 example), and concluded that none of them improved substantially on just assigning consecutive integers.

Cheers,
Bruce

Steve Simon, P.Mean Consulting wrote

I just noticed all the discussion about how you NEVER should compute the
mean of an ordinal variable. While I can't disagree with many of the
fine alternatives being suggested, I would mention that it is dangerous
to say "never".

Most people (including many statisticians) use a statistic called grade
point average (GPA) in their hiring decisions. This is clearly an
average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
to pretend that these numbers can be averaged? Well it implies that two
C's are equivalent to an A and an F. Still, the GPA doesn't do too much
violence to the data, and I think all of us would agree that a 3.5 GPA
is better than a 2.5 GPA.

Whether you should average or not depends a lot on the distribution of
the ordinal categories, the number of levels of the ordinal variable,
what the levels of the ordinal variable represent, and what the goals of
the statistical summary are.

I'd argue that the average of a 6 point variable is a simple summary
that is useful for descriptive statistics or informal comparisons. I'd
save the more complex alternatives for larger projects where a "quick
and dirty" approximation would be inappropriate.

I talk a bit more about this and point out the problems with a commonly
used alternative for ordinal data, rank-based methods, at my old website:
  * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
--
Steve Simon, Standard Disclaimer
Sign up for The Monthly Mean, the newsletter that
dares to call itself "average" at www.pmean.com/news
"Data entry and data management issues with examples
in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
Free webinar. Details at www.pmean.com/webinars

=====================
To manage your subscription to SPSSX-L, send a message to
[e-mail] (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

[Kornbrot, Diana]

Of course, averaging ordinal data is VERY common
It still a BAD idea, I am much more interested in the number of As/Number of courses taken than the mean GPA
I would certainly take someone with 2As, and 4Cs, gp = 16/6 = 2.67 or even 2As , 3 Cs and a D, gpa = 2.5, over someone with 6 Bs gpa =3
Using gpa encourages mediocrity over focussed excellence, and in my view feeds into passion and interest in knowledge in a very undesirable way.
The student with passionate interest in some topic has to sacrifice that passion to make sure of Bs in topics of peripheral interest. Is that good?

I do not recommend ranks which loses even more information than means.
I recommend proprotion or cumulative proportions whihch is more nearly aligned with what the decision maker needs to know.
This is equally true of applied problems such as customer satisfaction. One needs to know how many people hate as well as how many people like.
Averaging ordinal items lets the bad news get hidden in the average, as above a respectable average may be obtained while both the good and the bad news gets hidden.

If I am told that programme X has higher gpa than programme Y, I am unumpressed. I want to KNOW whether that is due to improvement of proportion of good or of weaker grades. For all we know there was a shift form p(d) > p ( c) associated with a slightly smaller shift form p(a) to p(b). Happy about that?

Its our job as folk with statistics knpwledge to discourage these practice, and if we cannot avoid giving the means, at lieast ensure that more appropriate (proportion measures) are also available, i.e calculateable from information provided

End of rant.....

Best

diana



On 09/08/2010 15:26, "Steve Simon, P.Mean Consulting" <net@...> wrote:

I just noticed all the discussion about how you NEVER should compute the
mean of an ordinal variable. While I can't disagree with many of the
fine alternatives being suggested, I would mention that it is dangerous
to say "never".

Most people (including many statisticians) use a statistic called grade
point average (GPA) in their hiring decisions. This is clearly an
average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
to pretend that these numbers can be averaged? Well it implies that two
C's are equivalent to an A and an F. Still, the GPA doesn't do too much
violence to the data, and I think all of us would agree that a 3.5 GPA
is better than a 2.5 GPA.

Whether you should average or not depends a lot on the distribution of
the ordinal categories, the number of levels of the ordinal variable,
what the levels of the ordinal variable represent, and what the goals of
the statistical summary are.

I'd argue that the average of a 6 point variable is a simple summary
that is useful for descriptive statistics or informal comparisons. I'd
save the more complex alternatives for larger projects where a "quick
and dirty" approximation would be inappropriate.

I talk a bit more about this and point out the problems with a commonly
used alternative for ordinal data, rank-based methods, at my old website:
  * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
--
Steve Simon, Standard Disclaimer
Sign up for The Monthly Mean, the newsletter that
dares to call itself "average" at www.pmean.com/news
"Data entry and data management issues with examples
in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
Free webinar. Details at www.pmean.com/webinars

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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

---

Professor Diana Kornbrot
email:  d.e.kornbrot@...    
web:    http://web.me.com/kornbrot/KornbrotHome.html
Work
Centre for Lifespan & Chronic Illness Research, CLiCIR
School of Psychology
University of Hertfordshire
College Lane, Hatfield, Hertfordshire AL10 9AB, UK
voice:  [phone]
Home
19 Elmhurst Avenue
London N2 0LT, UK
 voice:     [phone]
 mobile:   [phone]
fax:        [phone]

[John F Hall]

Re: 6 point variable
Why don't you just group the top two and bottom two, express them as % and
then subtract the bottom two from the top two and express that as % diff?
Most clients can understand that.  There are only two answers to any
question, "Yes" and "No", so my mates say, but at a pinch you can split
these with "Is that a little or a lot?" which gives you a compromise of four
points.

John Hall
http://surveyresearch.weebly.com


----- Original Message -----
From: Kornbrot, Diana
To: [hidden email]
Sent: Monday, August 09, 2010 5:25 PM
Subject: Re: 6 point variable


Of course, averaging ordinal data is VERY common
It still a BAD idea, I am much more interested in the number of As/Number of
courses taken than the mean GPA
I would certainly take someone with 2As, and 4Cs, gp = 16/6 = 2.67 or even
2As , 3 Cs and a D, gpa = 2.5, over someone with 6 Bs gpa =3
Using gpa encourages mediocrity over focussed excellence, and in my view
feeds into passion and interest in knowledge in a very undesirable way.
The student with passionate interest in some topic has to sacrifice that
passion to make sure of Bs in topics of peripheral interest. Is that good?

I do not recommend ranks which loses even more information than means.
I recommend proprotion or cumulative proportions whihch is more nearly
aligned with what the decision maker needs to know.
This is equally true of applied problems such as customer satisfaction. One
needs to know how many people hate as well as how many people like.
Averaging ordinal items lets the bad news get hidden in the average, as
above a respectable average may be obtained while both the good and the bad
news gets hidden.

If I am told that programme X has higher gpa than programme Y, I am
unumpressed. I want to KNOW whether that is due to improvement of proportion
of good or of weaker grades. For all we know there was a shift form p(d) > p
( c) associated with a slightly smaller shift form p(a) to p(b). Happy about
that?

Its our job as folk with statistics knpwledge to discourage these practice,
and if we cannot avoid giving the means, at lieast ensure that more
appropriate (proportion measures) are also available, i.e calculateable from
information provided

End of rant.....

Best

diana



On 09/08/2010 15:26, "Steve Simon, P.Mean Consulting" <[hidden email]> wrote:


I just noticed all the discussion about how you NEVER should compute the
mean of an ordinal variable. While I can't disagree with many of the
fine alternatives being suggested, I would mention that it is dangerous
to say "never".

Most people (including many statisticians) use a statistic called grade
point average (GPA) in their hiring decisions. This is clearly an
average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
to pretend that these numbers can be averaged? Well it implies that two
C's are equivalent to an A and an F. Still, the GPA doesn't do too much
violence to the data, and I think all of us would agree that a 3.5 GPA
is better than a 2.5 GPA.

Whether you should average or not depends a lot on the distribution of
the ordinal categories, the number of levels of the ordinal variable,
what the levels of the ordinal variable represent, and what the goals of
the statistical summary are.

I'd argue that the average of a 6 point variable is a simple summary
that is useful for descriptive statistics or informal comparisons. I'd
save the more complex alternatives for larger projects where a "quick
and dirty" approximation would be inappropriate.

I talk a bit more about this and point out the problems with a commonly
used alternative for ordinal data, rank-based methods, at my old website:
  * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
--
Steve Simon, Standard Disclaimer
Sign up for The Monthly Mean, the newsletter that
dares to call itself "average" at www.pmean.com/news
"Data entry and data management issues with examples
in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
Free webinar. Details at www.pmean.com/webinars

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






Professor Diana Kornbrot
email:  [hidden email]
web:    http://web.me.com/kornbrot/KornbrotHome.html
Work
Centre for Lifespan & Chronic Illness Research, CLiCIR
School of Psychology
University of Hertfordshire
College Lane, Hatfield, Hertfordshire AL10 9AB, UK
voice:  [phone]
Home
19 Elmhurst Avenue
London N2 0LT, UK
 voice:     [phone]
 mobile:   [phone]
fax:        [phone]

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

[Wilhelm Landerholm | Queue]

Hi!

So, John, 20% "Yes" and 20% "No" = 0% difference; and 45% "Yes" and 45% "No" = 0% difference.
But in your analysis they are treated the same? If that is the case I don't agree.

My recommendation is to keep information "simple" and do not forget that the distribution is at least as interesting as the average.

All the best

Wilhelm (Wille) Landerholm
[phone]

Queue/STATB 
Making sense of data
BOX 92 | 162 12 Vallingby | Sweden

2010/8/10 John F Hall <[hidden email]>

Re: 6 point variable
Why don't you just group the top two and bottom two, express them as % and
then subtract the bottom two from the top two and express that as % diff?
Most clients can understand that.  There are only two answers to any
question, "Yes" and "No", so my mates say, but at a pinch you can split
these with "Is that a little or a lot?" which gives you a compromise of four
points.

John Hall
http://surveyresearch.weebly.com

----- Original Message -----
From: Kornbrot, Diana
To: [hidden email]

Sent: Monday, August 09, 2010 5:25 PM
Subject: Re: 6 point variable

Of course, averaging ordinal data is VERY common
It still a BAD idea, I am much more interested in the number of As/Number of
courses taken than the mean GPA
I would certainly take someone with 2As, and 4Cs, gp = 16/6 = 2.67 or even
2As , 3 Cs and a D, gpa = 2.5, over someone with 6 Bs gpa =3
Using gpa encourages mediocrity over focussed excellence, and in my view
feeds into passion and interest in knowledge in a very undesirable way.
The student with passionate interest in some topic has to sacrifice that
passion to make sure of Bs in topics of peripheral interest. Is that good?

I do not recommend ranks which loses even more information than means.
I recommend proprotion or cumulative proportions whihch is more nearly
aligned with what the decision maker needs to know.
This is equally true of applied problems such as customer satisfaction. One
needs to know how many people hate as well as how many people like.
Averaging ordinal items lets the bad news get hidden in the average, as
above a respectable average may be obtained while both the good and the bad
news gets hidden.

If I am told that programme X has higher gpa than programme Y, I am
unumpressed. I want to KNOW whether that is due to improvement of proportion
of good or of weaker grades. For all we know there was a shift form p(d) > p
( c) associated with a slightly smaller shift form p(a) to p(b). Happy about
that?

Its our job as folk with statistics knpwledge to discourage these practice,
and if we cannot avoid giving the means, at lieast ensure that more
appropriate (proportion measures) are also available, i.e calculateable from
information provided

End of rant.....

Best

diana



On 09/08/2010 15:26, "Steve Simon, P.Mean Consulting" <[hidden email]> wrote:


I just noticed all the discussion about how you NEVER should compute the
mean of an ordinal variable. While I can't disagree with many of the
fine alternatives being suggested, I would mention that it is dangerous
to say "never".

Most people (including many statisticians) use a statistic called grade
point average (GPA) in their hiring decisions. This is clearly an
average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
to pretend that these numbers can be averaged? Well it implies that two
C's are equivalent to an A and an F. Still, the GPA doesn't do too much
violence to the data, and I think all of us would agree that a 3.5 GPA
is better than a 2.5 GPA.

Whether you should average or not depends a lot on the distribution of
the ordinal categories, the number of levels of the ordinal variable,
what the levels of the ordinal variable represent, and what the goals of
the statistical summary are.

I'd argue that the average of a 6 point variable is a simple summary
that is useful for descriptive statistics or informal comparisons. I'd
save the more complex alternatives for larger projects where a "quick
and dirty" approximation would be inappropriate.

I talk a bit more about this and point out the problems with a commonly
used alternative for ordinal data, rank-based methods, at my old website:
 * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
--
Steve Simon, Standard Disclaimer
Sign up for The Monthly Mean, the newsletter that
dares to call itself "average" at www.pmean.com/news
"Data entry and data management issues with examples
in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
Free webinar. Details at www.pmean.com/webinars

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






Professor Diana Kornbrot
email:  [hidden email]
web:    http://web.me.com/kornbrot/KornbrotHome.html
Work
Centre for Lifespan & Chronic Illness Research, CLiCIR
School of Psychology
University of Hertfordshire
College Lane, Hatfield, Hertfordshire AL10 9AB, UK
voice:  [phone]
Home
19 Elmhurst Avenue
London N2 0LT, UK
voice:     [phone]
mobile:   [phone]
fax:        [phone]

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

[Kornbrot, Diana]

Likert was very explicit about the difference between an item and a scale.
The scales of course sum ordinal variables.
As Bruce, states, Likert did look empirically at consequences of summing ordinal items. I believe he argues that if there are a ‘reasonable’ number of items then results based on scales are fairly robust, due to the central limit theorem, cf Dawes, R. M. (1979). The robust beauty of improper linear models American Psychologist , Vol. 34, pp. 571-582. (1610 citations!)

Still think NEVER is appropriate if extreme, for Likert ITEMS
Best
Diana

On 09/08/2010 16:17, "Bruce Weaver" <MailScanner has detected a possible fraud attempt from "bruce.weaver@..." claiming to be MailScanner has detected a possible fraud attempt from "bruce.weaver@..." claiming to be bruce.weaver@... <http://bruce.weaver@...> > wrote:

Nice note on your website, Steve.  I have two comments.

1. I believe that Likert would only have used the term "scale" after summing
or averaging over several items.  I.e., it takes more than one Likert-type
item to make a scale.

2. If I'm not mistaken, Likert tried several more complex methods of
assigning numbers to the ordinal categories (like your -3 -1 0 1 3 example),
and concluded that none of them improved substantially on just assigning
consecutive integers.

Cheers,
Bruce



Steve Simon, P.Mean Consulting wrote:
>
> I just noticed all the discussion about how you NEVER should compute the
> mean of an ordinal variable. While I can't disagree with many of the
> fine alternatives being suggested, I would mention that it is dangerous
> to say "never".
>
> Most people (including many statisticians) use a statistic called grade
> point average (GPA) in their hiring decisions. This is clearly an
> average computed on ordinal (A, B, C, D, F) data. Is it an awful thing
> to pretend that these numbers can be averaged? Well it implies that two
> C's are equivalent to an A and an F. Still, the GPA doesn't do too much
> violence to the data, and I think all of us would agree that a 3.5 GPA
> is better than a 2.5 GPA.
>
> Whether you should average or not depends a lot on the distribution of
> the ordinal categories, the number of levels of the ordinal variable,
> what the levels of the ordinal variable represent, and what the goals of
> the statistical summary are.
>
> I'd argue that the average of a 6 point variable is a simple summary
> that is useful for descriptive statistics or informal comparisons. I'd
> save the more complex alternatives for larger projects where a "quick
> and dirty" approximation would be inappropriate.
>
> I talk a bit more about this and point out the problems with a commonly
> used alternative for ordinal data, rank-based methods, at my old website:
>   * http://www.childrensmercy.org/stats/weblog2005/OrdinalData.asp
> --
> Steve Simon, Standard Disclaimer
> Sign up for The Monthly Mean, the newsletter that
> dares to call itself "average" at www.pmean.com/news
> "Data entry and data management issues with examples
> in IBM SPSS," Tuesday, August 24, 11am-noon CDT.
> Free webinar. Details at www.pmean.com/webinars
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> LISTSERV@... <http://LISTSERV@...>  (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@... <http://bweaver@...>
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/inexpensive-home-version-tp2259717p2268975.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... <http://LISTSERV@...>  (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

---

Professor Diana Kornbrot
email:  d.e.kornbrot@... <http://d.e.kornbrot@...>    
web:    http://web.me.com/kornbrot/KornbrotHome.html
Work
Centre for Lifespan & Chronic Illness Research, CLiCIR
School of Psychology
University of Hertfordshire
College Lane, Hatfield, Hertfordshire AL10 9AB, UK
voice:  [phone]
Home
19 Elmhurst Avenue
London N2 0LT, UK
 voice:     [phone]
 mobile:   [phone]
fax:        [phone]

[Richard Ristow]

At 02:04 PM 8/2/2010, Marta García-Granero wrote:

>C'mon, "us.ibm.com" people, say something...
>could we have SPSS 15 as a cheap home version?
>could we, could we, could we? pleeeease?

That is an interesting thought, which has also occurred to me:

The cheapest and easiest cut-down version SPSS
could offer, would be the regular system, several
releases back.  Some organizations effectively do
something like this, by buying only every few releases.

I think it'd have to be at least 4 releases back,
or too many full-pay commercial customers would
say 'what the heck', and get the cheap one.

With SPSS selling permanent licenses, as it does
(though not only those), users can move out of
SPSS with 'graceful degradation' of their
capabilities. That is, by simply staying with the
last version they have, users can migrate code
and files at whatever pace suits them -- their
SPSS will get less and less capable compared with
current versions, but won't disappear.

(By contrast, SAS's annual-license policy locks
you into their high annual fee essentially
forever: let a renewal go by, and EVERYTHING
stops working. Users who migrate must migrate
everything within one year, or else. That's one
reason small users almost never consider SAS.
And, SAS users: save EVERY file you care about,
in portable format, and make sure it can be read by something besides SAS.)

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

[Marta Garcia-Granero]

Hi Richard:

Richard Ristow wrote:
>
> That is an interesting thought, which has also occurred to me:
>
> The cheapest and easiest cut-down version SPSS could offer, would be
> the regular system, several releases back.  Some organizations
> effectively do something like this, by buying only every few releases.
>
> I think it'd have to be at least 4 releases back, or too many full-pay
> commercial customers would say 'what the heck', and get the cheap one.

In my opinion, that cut-down version of SPSS should be available only
for academic institutions, like Universities, to keep "full-pay
commercial customers" from benefiting of the cheap version. Thus, in the
short run, SPSS-IBM would get some money from customers they are going
to lose very soon (like my University), and, in the long run, SPSS would
still be known to graduated people, future employees of those companies
that would go on using the full version, instead of switching to another
one.

I hope someone up the chain of command in SPSS-IBM will get to read all
the exchange we've had concerning SPSS 15.

Best regards,
Marta GG


--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/

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

[Francien Berndsen]

> In my opinion, that cut-down version of SPSS should be available only
> for academic institutions, like Universities, to keep "full-pay
> commercial customers" from benefiting of the cheap version. Thus, in the
> short run, SPSS-IBM would get some money from customers they are going
> to lose very soon (like my University), and, in the long run, SPSS would
> still be known to graduated people, future employees of those companies
> that would go on using the full version, instead of switching to another
> one.

It's not only the academic world that's annoyed with the SPSS licensing
terms. My company had plans changing to Stata, but SPSS did us a better
offer at last.

When testing Stata, one of my main disadvantages was that importing and
exporting data to other programs (like excel/access) is far more difficult
then with SPSS. And we do that very often. You need some sort of converting
program.

Francien

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

[Nils Braakmann]

I always found outsheet/outfile and insheet in Stata to be fairly
powerful and relatively simple. (Of course, StatTransfer is way
easier.)

Cheers,
Nils

On Wed, Aug 18, 2010 at 7:05 AM, Francien Berndsen <[hidden email]> wrote:

>> In my opinion, that cut-down version of SPSS should be available only
>> for academic institutions, like Universities, to keep "full-pay
>> commercial customers" from benefiting of the cheap version. Thus, in the
>> short run, SPSS-IBM would get some money from customers they are going
>> to lose very soon (like my University), and, in the long run, SPSS would
>> still be known to graduated people, future employees of those companies
>> that would go on using the full version, instead of switching to another
>> one.
>
> It's not only the academic world that's annoyed with the SPSS licensing
> terms. My company had plans changing to Stata, but SPSS did us a better
> offer at last.
>
> When testing Stata, one of my main disadvantages was that importing and
> exporting data to other programs (like excel/access) is far more difficult
> then with SPSS. And we do that very often. You need some sort of converting
> program.
>
> Francien
>
> =====================
> 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

[SR Millis-3]

StatTransfer software does an exellent job of converting all types of databases:

http://www.stattransfer.com/

I have no financial interest in this software.


~~~~~~~~~~~
Scott R Millis, PhD, ABPP, CStat, CSci
Professor & Director of Research
Dept of Physical Medicine & Rehabilitation
Dept of Emergency Medicine
Wayne State University School of Medicine
261 Mack Blvd
Detroit, MI 48201
Email:  [hidden email]
Email:  [hidden email]
Tel: [phone]
Fax: [phone]


--- On Wed, 8/18/10, Francien Berndsen <[hidden email]> wrote:

> From: Francien Berndsen <[hidden email]>
> Subject: Re: inexpensive 'home' version?
> To: [hidden email]
> Date: Wednesday, August 18, 2010, 2:05 AM
> > In my opinion, that cut-down
> version of SPSS should be available only
> > for academic institutions, like Universities, to keep
> "full-pay
> > commercial customers" from benefiting of the cheap
> version. Thus, in the
> > short run, SPSS-IBM would get some money from
> customers they are going
> > to lose very soon (like my University), and, in the
> long run, SPSS would
> > still be known to graduated people, future employees
> of those companies
> > that would go on using the full version, instead of
> switching to another
> > one.
>
> It's not only the academic world that's annoyed with the
> SPSS licensing
> terms. My company had plans changing to Stata, but SPSS did
> us a better
> offer at last.
>
> When testing Stata, one of my main disadvantages was that
> importing and
> exporting data to other programs (like excel/access) is far
> more difficult
> then with SPSS. And we do that very often. You need some
> sort of converting
> program.
>
> Francien
>
> =====================
> 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