SPSSX Discussion

Eigenvalue greater than 1: Why it is the Default of SPSS?

Classic

List

Threaded

4 messages Options

Johnny Amora

Eigenvalue greater than 1: Why it is the Default of SPSS?

Hi all,

SPSS still used the "eigenvalues-greater-than-1" criterion as default for the selection of number of factors to retain in the FACTOR procedure inspite of the many studies reported that the eigenvalues-greater-than-1" criterion is a worst criterion. It was already proven that Parallel analysis is a recommended alternative to the worse eigenvalues-greater-than-1" criterion. But sad to note that SPSS does not provide the Parallel analysis procedure. I want to hear why SPSS is insensitive about this issue.

Johnny

How do you prepare for your first job interview?Discover the different ways in Yahoo! Answers!

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

Re: Eigenvalue greater than 1: Why it is the Default of SPSS?

Deciding on the number of factors to retain is very much a matter of
informed judgment. Technically one could extract as many factors as
there are variables.

One eigenvalue unit is "one variable's worth" of the total variance.
The assumption to stop extracting at that level is based on the idea
that one would not want any factors that had less. It is a rule of thumb
for *when to stop extracting*, not a rule of thumb for the* number to
retain*. "No way there are more factors needed" is not not the same as
"this is the number to retain".

Parallel analysis indeed gives an very good idea of where to start
looking for the number to retain. It is a ballpark.

"SPSS does not provide parallel analysis" is not the same as "one cannot
do parallel analysis using SPSS". Check the archives.

BTW parallel analysis can be be a very heavy load on many machines.

Art Kendall
Social Research Consultants

Johnny Amora wrote:

> Hi all,
>
> SPSS still used the "eigenvalues-greater-than-1" criterion as default for the selection of number of factors to retain in the FACTOR procedure inspite of the many studies reported that the eigenvalues-greater-than-1" criterion is a worst criterion. It was already proven that Parallel analysis is a recommended alternative to the worse eigenvalues-greater-than-1" criterion. But sad to note that SPSS does not provide the Parallel analysis procedure. I want to hear why SPSS is insensitive about this issue.
>
> Johnny
>
>
> How do you prepare for your first job interview?Discover the different ways in Yahoo! Answers!
>
> ===================
> 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
Social Research Consultants

SPSS Support

Re: Eigenvalue greater than 1: Why it is the Default of SPSS?

Hello Johnny, Art, and others,
A feature request to add parallel analysis and Velicer's MAP to the Factor procedure has been filed with SPSS Development. I've pasted a related resolution from the SPSS Support web site ( http://support.spss.com ) below. It contains a link to Brian O'Connor's site with SPSS syntax for parallel analysis and Velicer's MAP test.

David Matheson
SPSS Statistical Support
********

Resolution number: 76126 Created on: Mar 3 2008 Last Reviewed on: Mar 3 2008 Resolution Status: Published - External

Problem Subject: FACTOR enhancement request: Parallel analysis and MAP test to help determine number of factors

Problem Description: Does SPSS FACTOR offer parallel analysis or Velicer's MAP test for choosing the number of factors to retain?

Resolution Subject: SPSS FACTOR does not currently offer parallel analysis or the MAP test

Resolution Description:
SPSS does not currently offer either of these methods. An enhancement request has been filed with SPSS Development. SPSS command syntax for these methods is available from http://people.ok.ubc.ca/brioconn/nfactors/nfactors.html . Please note that this is an external web site and SPSS cannot provide support for these commands or take responsibility for the results produced.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
Sent: Tuesday, November 25, 2008 11:05 AM
To: [hidden email]
Subject: Re: Eigenvalue greater than 1: Why it is the Default of SPSS?

Deciding on the number of factors to retain is very much a matter of informed judgment. Technically one could extract as many factors as there are variables.

One eigenvalue unit is "one variable's worth" of the total variance.
The assumption to stop extracting at that level is based on the idea that one would not want any factors that had less. It is a rule of thumb for *when to stop extracting*, not a rule of thumb for the* number to
retain*. "No way there are more factors needed" is not not the same as
"this is the number to retain".

Parallel analysis indeed gives an very good idea of where to start looking for the number to retain. It is a ballpark.

"SPSS does not provide parallel analysis" is not the same as "one cannot do parallel analysis using SPSS". Check the archives.

BTW parallel analysis can be be a very heavy load on many machines.

Art Kendall
Social Research Consultants

Johnny Amora wrote:

> Hi all,
>
> SPSS still used the "eigenvalues-greater-than-1" criterion as default for the selection of number of factors to retain in the FACTOR procedure inspite of the many studies reported that the eigenvalues-greater-than-1" criterion is a worst criterion. It was already proven that Parallel analysis is a recommended alternative to the worse eigenvalues-greater-than-1" criterion. But sad to note that SPSS does not provide the Parallel analysis procedure. I want to hear why SPSS is insensitive about this issue.
>
> Johnny
>
>
> How do you prepare for your first job interview?Discover the different ways in Yahoo! Answers!
>
> ===================
> 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
>
>
>

Jims More

Re: Eigenvalue greater than 1: Why it is the Default of SPSS?

Hi all,

If I may add, I know SPSS upgrades every year their products, but, aside from Parallel Analysis, I dont know why they dont have "direct procedure" for polychoric/tetrachoric correlation. Hope to include it in the next version because polychoric/tetrachoric correlation matrix is needed for their AMOS as input data.

Jims

--- On Tue, 25/11/08, SPSS Support <[hidden email]> wrote:

From: SPSS Support <[hidden email]>
Subject: Re: Eigenvalue greater than 1: Why it is the Default of SPSS?
To: [hidden email]
Received: Tuesday, 25 November, 2008, 10:21 PM

Hello Johnny, Art, and others,
A feature request to add parallel analysis and Velicer's MAP to the Factor
procedure has been filed with SPSS Development. I've pasted a related
resolution from the SPSS Support web site ( http://support.spss.com ) below.. It
contains a link to Brian O'Connor's site with SPSS syntax for parallel
analysis and Velicer's MAP test.

David Matheson
SPSS Statistical Support
********

Resolution number: 76126 Created on: Mar 3 2008 Last Reviewed on: Mar 3 2008
Resolution Status: Published - External

Problem Subject: FACTOR enhancement request: Parallel analysis and MAP test to
help determine number of factors

Problem Description: Does SPSS FACTOR offer parallel analysis or Velicer's
MAP test for choosing the number of factors to retain?

Resolution Subject: SPSS FACTOR does not currently offer parallel analysis or
the MAP test

Resolution Description:
SPSS does not currently offer either of these methods. An enhancement request
has been filed with SPSS Development. SPSS command syntax for these methods is
available from http://people.ok.ubc.ca/brioconn/nfactors/nfactors.html . Please
note that this is an external web site and SPSS cannot provide support for these
commands or take responsibility for the results produced.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art
Kendall
Sent: Tuesday, November 25, 2008 11:05 AM
To: [hidden email]
Subject: Re: Eigenvalue greater than 1: Why it is the Default of SPSS?

Deciding on the number of factors to retain is very much a matter of informed
judgment. Technically one could extract as many factors as there are variables.

One eigenvalue unit is "one variable's worth" of the total
variance.
The assumption to stop extracting at that level is based on the idea that one
would not want any factors that had less. It is a rule of thumb for *when to
stop extracting*, not a rule of thumb for the* number to
retain*. "No way there are more factors needed" is not not the same
as
"this is the number to retain".

Parallel analysis indeed gives an very good idea of where to start looking for
the number to retain. It is a ballpark.

"SPSS does not provide parallel analysis" is not the same as
"one cannot do parallel analysis using SPSS". Check the archives.

BTW parallel analysis can be be a very heavy load on many machines.

Art Kendall
Social Research Consultants

Johnny Amora wrote:
> Hi all,
>
> SPSS still used the "eigenvalues-greater-than-1" criterion as
default for the selection of number of factors to retain in the FACTOR procedure
inspite of the many studies reported that the eigenvalues-greater-than-1"
criterion is a worst criterion. It was already proven that Parallel analysis is
a recommended alternative to the worse eigenvalues-greater-than-1"
criterion. But sad to note that SPSS does not provide the Parallel analysis
procedure. I want to hear why SPSS is insensitive about this issue.
>
> Johnny
>
>
> How do you prepare for your first job interview?Discover the
different ways in Yahoo! Answers!

>
> ===================
> 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

Start your day with Yahoo!7 and win a Sony Bravia TV. Enter now http://au.docs.yahoo.com/homepageset/?p1=other&p2=au&p3=tagline

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