Unequal n question

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

Unequal n question

Salbod
Good Day, Dear Friends,

Your help on my missing data problem was priceless. You gave a welcomed
stack of references which I shall enjoy this summer.

I had another problem that I thought my office stat books could answer, but
I was wrong.

I have two variables that I want to correlated in non-experimental design.
One variable called parenting style is ordered on continuous scale. The
other variable called parenting type is a categorical measure, where
participates select their parenting type out of three descriptions; note,
the types are like apples, oranges, and plums, i.e., they can't be combined.
The ns that I obtained for type were 4, 180, and 20. Can I use a one-way (
regression with dummy codes) with these size ns?

I remembered reading years ago that I could drop the first group because it
had a ratio 1 to 45 with the second group.  I can't find the reference. Did
I dream it?

Any suggestions or references are  most welcome.

TIA

Your friend,

Stephen Salbod, Pace University, NYC
Reply | Threaded
Open this post in threaded view
|

Re: Unequal n question

Hector Maletta
         Stephen,
         What matters most is not the ratio of ns of the various categories,
but the fact that one of them is populated by only 4 cases. The kind of
analysis you might want to conduct, such as ANOVA for instance, requires a
sufficient number of cases in each group to assess within-group variance,
which is not reasonably measurable with only 4 cases.
         Classical ANOVA, indeed, was originally designed for groups of
equal size, but adjustments exist for designs with unequal sizes, and SPSS
can do that. If your groups A and B were of sizes 40,000 and 1,200,000
instead of 4 and 120, an ANOVA with unequal group sizes could be conducted,
but 4 cases is definitely too few.
         You might conceivably contrast the second group against the rest
(the sum of the other two groups, n=24), but I wonder whether that is
substantively useful.
         Apparently your parental styles were defined in such a way that
most cases fell in only one big category. In other groups, the parental
style variable has low "discriminative power": it saw no differences among
120 cases, and isolated only 4 cases on one side and 20 on the other. One
should want a more equilibrated categorization, perhaps.
         Hector

         -----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Stephen Salbod
Sent: 29 June 2007 16:18
To: [hidden email]
Subject: Unequal n question

         Good Day, Dear Friends,

         Your help on my missing data problem was priceless. You gave a
welcomed
         stack of references which I shall enjoy this summer.

         I had another problem that I thought my office stat books could
answer, but
         I was wrong.

         I have two variables that I want to correlated in non-experimental
design.
         One variable called parenting style is ordered on continuous scale.
The
         other variable called parenting type is a categorical measure,
where
         participates select their parenting type out of three descriptions;
note,
         the types are like apples, oranges, and plums, i.e., they can't be
combined.
         The ns that I obtained for type were 4, 180, and 20. Can I use a
one-way (
         regression with dummy codes) with these size ns?

         I remembered reading years ago that I could drop the first group
because it
         had a ratio 1 to 45 with the second group.  I can't find the
reference. Did
         I dream it?

         Any suggestions or references are  most welcome.

         TIA

         Your friend,

         Stephen Salbod, Pace University, NYC
Reply | Threaded
Open this post in threaded view
|

Visual Studio 2005 VB call SPSS

paul yeo
Dear all

can any one tell me how to use the above to call SPSS
?
in the pass i use VB6 , and i add SPSS type lib , it
work , now i change to Visual Stuidio 2005 VB , i use
back the same lib , it does not work.

can some one tell me where can i download the lib ?

thank

paul




____________________________________________________
Yahoo! Singapore Answers
Real people. Real questions. Real answers. Share what you know at http://answers.yahoo.com.sg
Reply | Threaded
Open this post in threaded view
|

Lib for machine learning...

Nivruti Hinge
Hi,

Can anyone suggest any open source lib for machine learning which can be
called from SPSS using python?


-Regards,
Nivruti Hinge

Module Lead
Persistent Systems Pvt. Ltd.
http://www.persistentsys.com
Direct : +91 (20) 30235627

____________________________________________________________________________
_______________
Persistent Systems - Software Development Partner for Competitive Advantage.
Persistent Systems provides custom software product development services.
With over 15 years, 140 customers, and 700 release cycles experience, we
deliver unmatched value of higher quality, faster time to market and lower
total costs.



DISCLAIMER
==========
This e-mail may contain privileged and confidential information which is the property of Persistent Systems Pvt. Ltd. It is intended only for the use of the individual or entity to which it is addressed. If you are not the intended recipient, you are not authorized to read, retain, copy, print, distribute or use this message. If you have received this communication in error, please notify the sender and delete all copies of this message. Persistent Systems Pvt. Ltd. does not accept any liability for virus infected mails.
Reply | Threaded
Open this post in threaded view
|

Re: Lib for machine learning...

Fabien Couprie
Hi you should try Orange on : http://www.ailab.si/orange/
Beware as usual with free datamining tools, the treatment are executed fully
in RAM. So it is not the best way to treat big datafiles. I never used it
myself, but it is enterely build with the beautiful Python language.  You
could try the datamining R packages, there is two Python bridge here
http://rpy.sourceforge.net/, http://www.omegahat.org/RSPython/.
Whatever the solution you choose i assume it won't be easy except if you are
a Python's God. I'm very interested by the subject, so if you please let me
know.


---------------------------------------------------------------
Fabien COUPRIE
Responsable recherches et études statistiques
email : [hidden email]
---------------------------------------------------------------
GOLDEN EYES SA
139, rue des Arts BP 70147 | 59 053 Roubaix Cedex 1
Tel. 03 59 57 70 00 | Fax 03 59 57 70 01
Web. www.golden-eyes.com
---------------------------------------------------------------
Reply | Threaded
Open this post in threaded view
|

Re: Visual Studio 2005 VB call SPSS

John McConnell
In reply to this post by paul yeo
Dear Paul

The same library should work.

You may need to add the library as a reference again inside VS2005 (it
is a COM library still rather than a .NET)

hth

John

John McConnell
applied insights
7200 The Quorum
Oxford Business Park North
Garsington Road
Oxford, OX4 2JZ

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
paul yeo
Sent: 02 July 2007 06:41
To: [hidden email]
Subject: Visual Studio 2005 VB call SPSS

Dear all

can any one tell me how to use the above to call SPSS
?
in the pass i use VB6 , and i add SPSS type lib , it
work , now i change to Visual Stuidio 2005 VB , i use
back the same lib , it does not work.

can some one tell me where can i download the lib ?

thank

paul




____________________________________________________
Yahoo! Singapore Answers
Real people. Real questions. Real answers. Share what you know at
http://answers.yahoo.com.sg
Reply | Threaded
Open this post in threaded view
|

Re: Lib for machine learning...

Art Kendall
In reply to this post by Nivruti Hinge
Much of data mining can be done within spss.  It may not be necessary to
go outside the main package.

supervised learning in statistical terminology is often discriminant
function analysis

unsupervised learning in statistical terminology is often clustering or
trees.

feature selection in statistical terminology is often "stepwise"

data reduction in statistical terminology is often factor analysis (PCA,
PFA, CATPCA).

Art Kendall
Social Research Consultants

Nivruti Hinge wrote:

> Hi,
>
> Can anyone suggest any open source lib for machine learning which can be
> called from SPSS using python?
>
>
> -Regards,
> Nivruti Hinge
>
> Module Lead
> Persistent Systems Pvt. Ltd.
> http://www.persistentsys.com
> Direct : +91 (20) 30235627
>
> ____________________________________________________________________________
> _______________
> Persistent Systems - Software Development Partner for Competitive Advantage.
> Persistent Systems provides custom software product development services.
> With over 15 years, 140 customers, and 700 release cycles experience, we
> deliver unmatched value of higher quality, faster time to market and lower
> total costs.
>
>
>
> DISCLAIMER
> ==========
> This e-mail may contain privileged and confidential information which is the property of Persistent Systems Pvt. Ltd. It is intended only for the use of the individual or entity to which it is addressed. If you are not the intended recipient, you are not authorized to read, retain, copy, print, distribute or use this message. If you have received this communication in error, please notify the sender and delete all copies of this message. Persistent Systems Pvt. Ltd. does not accept any liability for virus infected mails.
>
>
>
Art Kendall
Social Research Consultants
Reply | Threaded
Open this post in threaded view
|

How to set SPSS ver13 work like Ver 9 ?

paul yeo
In reply to this post by John McConnell
Dear all,

Thank you john for reply my email . i will try to
modify the VB code hope it will run well.

by the way i have one more questions , in the pass i
use VB6 to work with SPSS . for data analysis. it work
very well for last ten year.

when i try to test with SPSS ver13 , i notice that it
does not work well, the problem is each time i use VB
to run the report , it keep open new data editor , and
also keep open SPSS viewer.

end up i run out of memory.

can any one teach me how to set , so that SPSS ver 13
will behavious like SPSS9 ?

thank you for your reply.

paul yeo






Paul Yeo

--- John McConnell <[hidden email]>
wrote:

>
> Dear Paul
>
> The same library should work.
>
> You may need to add the library as a reference again
> inside VS2005 (it
> is a COM library still rather than a .NET)
>
> hth
>
> John
>
> John McConnell
> applied insights
> 7200 The Quorum
> Oxford Business Park North
> Garsington Road
> Oxford, OX4 2JZ
>
> -----Original Message-----
> From: SPSSX(r) Discussion
> [mailto:[hidden email]] On Behalf Of
> paul yeo
> Sent: 02 July 2007 06:41
> To: [hidden email]
> Subject: Visual Studio 2005 VB call SPSS
>
> Dear all
>
> can any one tell me how to use the above to call
> SPSS
> ?
> in the pass i use VB6 , and i add SPSS type lib , it
> work , now i change to Visual Stuidio 2005 VB , i
> use
> back the same lib , it does not work.
>
> can some one tell me where can i download the lib ?
>
> thank
>
> paul
>
>
>
>
> ____________________________________________________
> Yahoo! Singapore Answers
> Real people. Real questions. Real answers. Share
> what you know at
> http://answers.yahoo.com.sg
>
>
>




____________________________________________________
Yahoo! Singapore Answers
Real people. Real questions. Real answers. Share what you know at http://answers.yahoo.com.sg
Reply | Threaded
Open this post in threaded view
|

Re: How to set SPSS ver13 work like Ver 9 ?

John McConnell
Dear Paul

We might need to see the first few lines of code to understand exactly
why - it could be around handling multiple data documents/objects which
is a new feature in SPSS (I think it came in 13, maybe later) - but
allow me to guess a little.

If you want to open a new SPSS session (or "create a new SPSS Instance"
in object speak) then I imagine you will use either...

Dim objSPSS As spsswin.Application
objSPSS = createobject("spss.application")

...Or something like...
Dim objSPSS As new spsswin.Application

...in either case using ...
objSPSS.quit
... at the end of your application "should" close that instance of SPSS

Alternatively, if you already have an SPSS session open (i.e. an
instance available) you could just connect to that using ...

Dim objSPSS As spsswin.Application
objSPSS = GetObject(, "spss.application")

There is slightly smarter code which can connect to an instance if one
exists and create one if none exists. If the above doesn't help ... and
you think that would be helpful I can forward you another example.

hth

john

-----Original Message-----
From: paul yeo [mailto:[hidden email]]
Sent: 03 July 2007 09:52
To: John McConnell; [hidden email]
Subject: How to set SPSS ver13 work like Ver 9 ?

Dear all,

Thank you john for reply my email . i will try to
modify the VB code hope it will run well.

by the way i have one more questions , in the pass i
use VB6 to work with SPSS . for data analysis. it work
very well for last ten year.

when i try to test with SPSS ver13 , i notice that it
does not work well, the problem is each time i use VB
to run the report , it keep open new data editor , and
also keep open SPSS viewer.

end up i run out of memory.

can any one teach me how to set , so that SPSS ver 13
will behavious like SPSS9 ?

thank you for your reply.

paul yeo






Paul Yeo

--- John McConnell <[hidden email]>
wrote:

>
> Dear Paul
>
> The same library should work.
>
> You may need to add the library as a reference again
> inside VS2005 (it
> is a COM library still rather than a .NET)
>
> hth
>
> John
>
> John McConnell
> applied insights
> 7200 The Quorum
> Oxford Business Park North
> Garsington Road
> Oxford, OX4 2JZ
>
> -----Original Message-----
> From: SPSSX(r) Discussion
> [mailto:[hidden email]] On Behalf Of
> paul yeo
> Sent: 02 July 2007 06:41
> To: [hidden email]
> Subject: Visual Studio 2005 VB call SPSS
>
> Dear all
>
> can any one tell me how to use the above to call
> SPSS
> ?
> in the pass i use VB6 , and i add SPSS type lib , it
> work , now i change to Visual Stuidio 2005 VB , i
> use
> back the same lib , it does not work.
>
> can some one tell me where can i download the lib ?
>
> thank
>
> paul
>
>
>
>
> ____________________________________________________
> Yahoo! Singapore Answers
> Real people. Real questions. Real answers. Share
> what you know at
> http://answers.yahoo.com.sg
>
>
>




____________________________________________________
Yahoo! Singapore Answers
Real people. Real questions. Real answers. Share what you know at
http://answers.yahoo.com.sg
Reply | Threaded
Open this post in threaded view
|

Re: How to set SPSS ver13 work like Ver 9 ?

paul yeo
Dear john

your guess is correct , i will try to use quite
command to colse spss file .

paul yeo

--- John McConnell <[hidden email]>
wrote:

>
> Dear Paul
>
> We might need to see the first few lines of code to
> understand exactly
> why - it could be around handling multiple data
> documents/objects which
> is a new feature in SPSS (I think it came in 13,
> maybe later) - but
> allow me to guess a little.
>
> If you want to open a new SPSS session (or "create a
> new SPSS Instance"
> in object speak) then I imagine you will use
> either...
>
> Dim objSPSS As spsswin.Application
> objSPSS = createobject("spss.application")
>
> ...Or something like...
> Dim objSPSS As new spsswin.Application
>
> ...in either case using ...
> objSPSS.quit
> ... at the end of your application "should" close
> that instance of SPSS
>
> Alternatively, if you already have an SPSS session
> open (i.e. an
> instance available) you could just connect to that
> using ...
>
> Dim objSPSS As spsswin.Application
> objSPSS = GetObject(, "spss.application")
>
> There is slightly smarter code which can connect to
> an instance if one
> exists and create one if none exists. If the above
> doesn't help ... and
> you think that would be helpful I can forward you
> another example.
>
> hth
>
> john
>
> -----Original Message-----
> From: paul yeo [mailto:[hidden email]]
> Sent: 03 July 2007 09:52
> To: John McConnell; [hidden email]
> Subject: How to set SPSS ver13 work like Ver 9 ?
>
> Dear all,
>
> Thank you john for reply my email . i will try to
> modify the VB code hope it will run well.
>
> by the way i have one more questions , in the pass i
> use VB6 to work with SPSS . for data analysis. it
> work
> very well for last ten year.
>
> when i try to test with SPSS ver13 , i notice that
> it
> does not work well, the problem is each time i use
> VB
> to run the report , it keep open new data editor ,
> and
> also keep open SPSS viewer.
>
> end up i run out of memory.
>
> can any one teach me how to set , so that SPSS ver
> 13
> will behavious like SPSS9 ?
>
> thank you for your reply.
>
> paul yeo
>
>
>
>
>
>
> Paul Yeo
>
> --- John McConnell
> <[hidden email]>
> wrote:
>
> >
> > Dear Paul
> >
> > The same library should work.
> >
> > You may need to add the library as a reference
> again
> > inside VS2005 (it
> > is a COM library still rather than a .NET)
> >
> > hth
> >
> > John
> >
> > John McConnell
> > applied insights
> > 7200 The Quorum
> > Oxford Business Park North
> > Garsington Road
> > Oxford, OX4 2JZ
> >
> > -----Original Message-----
> > From: SPSSX(r) Discussion
> > [mailto:[hidden email]] On Behalf Of
> > paul yeo
> > Sent: 02 July 2007 06:41
> > To: [hidden email]
> > Subject: Visual Studio 2005 VB call SPSS
> >
> > Dear all
> >
> > can any one tell me how to use the above to call
> > SPSS
> > ?
> > in the pass i use VB6 , and i add SPSS type lib ,
> it
> > work , now i change to Visual Stuidio 2005 VB , i
> > use
> > back the same lib , it does not work.
> >
> > can some one tell me where can i download the lib
> ?
> >
> > thank
> >
> > paul
> >
> >
> >
> >
> >
> ____________________________________________________
> > Yahoo! Singapore Answers
> > Real people. Real questions. Real answers. Share
> > what you know at
> > http://answers.yahoo.com.sg
> >
> >
> >
>
>
>
>
> ____________________________________________________
>
> Yahoo! Singapore Answers
> Real people. Real questions. Real answers. Share
> what you know at
> http://answers.yahoo.com.sg
>
>
>




____________________________________________________
Yahoo! Singapore Answers
Real people. Real questions. Real answers. Share what you know at http://answers.yahoo.com.sg