SPSSX Discussion

HLM-SPSS Tutorial

Classic

List

Threaded

10 messages Options

Kizzy M Parks

HLM-SPSS Tutorial

Hello:
Does anyone know of an on-line HLM-SPSS tutorial?
If so, please reply with the link.

Thanks!

--
Kizzy M. Parks, Ph.D

Florida Tech

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

vlad simion

Re: HLM-SPSS Tutorial

Hi Kizzy,

try this links:

http://www.ats.ucla.edu/stat/
http://www.ssicentral.com/hlm/examples.html

Hth,
Vlad

On Thu, Jul 10, 2008 at 9:22 AM, Kizzy M Parks <[hidden email]> wrote:

> Hello:
> Does anyone know of an on-line HLM-SPSS tutorial?
> If so, please reply with the link.
>
> Thanks!
>
> --
> Kizzy M. Parks, Ph.D
>
> Florida Tech
>
> =====================
> 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
>

Neda Faregh

Re: HLM-SPSS Tutorial

Kizzy,
I don't know of an online HLM tutorial for SPSS, but would caution you
against using HLM in SPSS.
I have done HLM analyses in recent past first with SPSS and then with HLM 6
softwre.
The results are different. This difference is not minor either, you will
draw different conclusions (opposite in my case) from the results of each
software.
I do prefer HLM 6 and believe it to be the one that gives correct answers.
There are a couple of books that deal with HLM in SPSS specifically. One is
Leech, Barrett, and Morgan's SPSS for Intermediate statistics. The other is
of course Norusis' SPSS 16.0 advanced statistical procedures companion.
Of the two I much prefer Norusis' book.

Neda
Carleton U.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
vlad simion
Sent: Thursday, July 10, 2008 2:55 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Hi Kizzy,

try this links:

http://www.ats.ucla.edu/stat/
http://www.ssicentral.com/hlm/examples.html

Hth,
Vlad

On Thu, Jul 10, 2008 at 9:22 AM, Kizzy M Parks <[hidden email]> wrote:

Stevan Nielsen

Re: HLM-SPSS Tutorial (Doing HLM in SPSS)

To All, especially to the publishers of SPSS,

On a related note: Would it not be quite helpful and not all that complicated to expand the SPSS Aggregate function to compute regression components?

It seems to me that if Aggregate included Intercept and Regression Coefficient functions it would be possible to do rough and ready, but easy to control and examine, hierarchical modeling based on the products of Aggregate runs.

(By the way, Aggregate gets my vote as the most useful recent addition to SPSS functionality.)

Stevan Lars Nielsen, Ph.D.
Clinical Professor
Clinical Psychologist
Counseling and Career Center
Brigham Young University

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Neda Faregh
Sent: Thursday, July 10, 2008 6:25 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Kizzy,
I don't know of an online HLM tutorial for SPSS, but would caution you
against using HLM in SPSS.
I have done HLM analyses in recent past first with SPSS and then with HLM 6
softwre.
The results are different. This difference is not minor either, you will
draw different conclusions (opposite in my case) from the results of each
software.
I do prefer HLM 6 and believe it to be the one that gives correct answers.
There are a couple of books that deal with HLM in SPSS specifically. One is
Leech, Barrett, and Morgan's SPSS for Intermediate statistics. The other is
of course Norusis' SPSS 16.0 advanced statistical procedures companion.
Of the two I much prefer Norusis' book.

Neda
Carleton U.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
vlad simion
Sent: Thursday, July 10, 2008 2:55 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Hi Kizzy,

try this links:

http://www.ats.ucla.edu/stat/
http://www.ssicentral.com/hlm/examples.html

Hth,
Vlad

On Thu, Jul 10, 2008 at 9:22 AM, Kizzy M Parks <[hidden email]> wrote:

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

Peck, Jon

Re: HLM-SPSS Tutorial (Doing HLM in SPSS)

Why not use SPLIT FILES and OMS with sav format for this? That would work with (almost) any SPSS procedure to give you a dataset of resuilts for each split?

Regards,
Jon Peck

-----Original Message-----
From: SPSSX(r) Discussion <[hidden email]>
To: [hidden email] <[hidden email]>
Sent: Thu Jul 10 10:32:13 2008
Subject: Re: [SPSSX-L] HLM-SPSS Tutorial (Doing HLM in SPSS)

To All, especially to the publishers of SPSS,

On a related note: Would it not be quite helpful and not all that complicated to expand the SPSS Aggregate function to compute regression components?

It seems to me that if Aggregate included Intercept and Regression Coefficient functions it would be possible to do rough and ready, but easy to control and examine, hierarchical modeling based on the products of Aggregate runs.

(By the way, Aggregate gets my vote as the most useful recent addition to SPSS functionality.)

Stevan Lars Nielsen, Ph.D.
Clinical Professor
Clinical Psychologist
Counseling and Career Center
Brigham Young University

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Neda Faregh
Sent: Thursday, July 10, 2008 6:25 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Kizzy,
I don't know of an online HLM tutorial for SPSS, but would caution you
against using HLM in SPSS.
I have done HLM analyses in recent past first with SPSS and then with HLM 6
softwre.
The results are different. This difference is not minor either, you will
draw different conclusions (opposite in my case) from the results of each
software.
I do prefer HLM 6 and believe it to be the one that gives correct answers.
There are a couple of books that deal with HLM in SPSS specifically. One is
Leech, Barrett, and Morgan's SPSS for Intermediate statistics. The other is
of course Norusis' SPSS 16.0 advanced statistical procedures companion.
Of the two I much prefer Norusis' book.

Neda
Carleton U.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
vlad simion
Sent: Thursday, July 10, 2008 2:55 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Hi Kizzy,

try this links:

http://www.ats.ucla.edu/stat/
http://www.ssicentral.com/hlm/examples.html

Hth,
Vlad

On Thu, Jul 10, 2008 at 9:22 AM, Kizzy M Parks <[hidden email]> wrote:

Richard Ristow

Re: HLM-SPSS Tutorial (Doing HLM in SPSS)

In reply to this post by Stevan Nielsen

At 11:32 AM 7/10/2008, Stevan Nielsen wrote:

>(By the way, Aggregate gets my vote as the most useful recent
>addition to SPSS functionality.)

(Grin) AGGREGATE is immensely useful; I've posted more solutions
using AGGREGATE than I'd care to count. AGGREGATE is a main reason
SPSS works so well with files with 'long' data organization, usually
more easily than those with 'wide' organization.

But it's not a recent addition. It's been in SPSS at least since
release 6 for the mainframe (NOT SPSS 6 for Windows!), in 1975.

AGGREGATE came into its own when MATCH FILES was added, in SPSSX
[release 1] for mainframe and Vax, c. 1983. And it's had excellent
extensions since: MODE=ADDVARIABLES; MEDIAN as an aggregation
function; the current highly efficient implementation that does not
require the data be sorted.

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

Dennis Deck

Re: HLM-SPSS Tutorial

In reply to this post by Kizzy M Parks

The alogroithms used in various HLM packages differ somewhat so some
variation in results is to be expected. That is certainly true of MIXED
and HLM. But you must also ensure that your code specified analyses
that were truly identical. I ran a test of SPSS MIXED and HLM some time
ago and came up with nearly identical results (to several decimal
places) for that particular data set.

Most of my analyses, however, involve binary outcomes - unfortunately
SPSS MIXED cannot handle binary outcomes.

Dennis Deck, PhD
RMC Research Corporation
111 SW Columbia Street, Suite 1200
Portland, Oregon 97201-5843
voice: 503-223-8248 x715
voice: 800-788-1887 x715
fax: 503-223-8248
[hidden email]

-----Original Message-----
From: Neda Faregh [mailto:[hidden email]]
Sent: Thursday, July 10, 2008 5:25 AM
Subject: Re: HLM-SPSS Tutorial

Kizzy,
I don't know of an online HLM tutorial for SPSS, but would caution you
against using HLM in SPSS.
I have done HLM analyses in recent past first with SPSS and then with
HLM 6 softwre.
The results are different. This difference is not minor either, you will
draw different conclusions (opposite in my case) from the results of
each software.
I do prefer HLM 6 and believe it to be the one that gives correct
answers.
There are a couple of books that deal with HLM in SPSS specifically. One
is Leech, Barrett, and Morgan's SPSS for Intermediate statistics. The
other is of course Norusis' SPSS 16.0 advanced statistical procedures
companion.
Of the two I much prefer Norusis' book.

Neda
Carleton U.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
vlad simion
Sent: Thursday, July 10, 2008 2:55 AM
To: [hidden email]
Subject: Re: HLM-SPSS Tutorial

Hi Kizzy,

try this links:

http://www.ats.ucla.edu/stat/
http://www.ssicentral.com/hlm/examples.html

Hth,
Vlad

On Thu, Jul 10, 2008 at 9:22 AM, Kizzy M Parks <[hidden email]> wrote:

> Hello:
> Does anyone know of an on-line HLM-SPSS tutorial?
> If so, please reply with the link.
>
> Thanks!
>
> --
> Kizzy M. Parks, Ph.D
>
> Florida Tech
>
> =====================
> 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

Chen Wei-2

Heteroscedacity

In reply to this post by Neda Faregh

Dear All,

Could anyone provide any materials on heteroscedacity issue, how to detect
it and how to deal with it. Thanks very much.

Wayne

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

lts1

Re: Heteroscedacity

I would recommend Gwilym Pryce's Heteroscedasticity Testing and Correcting in SPSS. You can find it on Raynald's web site:

http://www.spsstools.net/spss.htm#SpecializedTutorials

Best,
Lisa

Lisa T. Stickney, Ph.D.
Assistant Professor
Merrick School of Business
U. of Baltimore

----- Original Message -----
From: "Chen Wei" <[hidden email]>
To: <[hidden email]>
Sent: Saturday, July 12, 2008 9:39 PM
Subject: Heteroscedacity

> Dear All,
>
> Could anyone provide any materials on heteroscedacity issue, how to detect
> it and how to deal with it. Thanks very much.
>
> Wayne
>
> =====================
> 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 Hitchin

Re: Heteroscedacity

Quoting Lisa Stickney <[hidden email]>:

> I would recommend Gwilym Pryce's Heteroscedasticity Testing and >
Correcting in SPSS. You can find it on Raynald's web site:
>
> http://www.spsstools.net/spss.htm#SpecializedTutorials
>
That's a very useful reference.

It may also be worth considering that there are some problems which look
like heteroscedasticity but which have a more fundamental reason.

One is the inappropriate choice of method for a variable which is
inherently non-normal. When events are counted the result is inevitably
non-normal (because counts are whole numbers rather than continuous and
can't be negative for a start). Such variables which have variances
which are (roughly) equal to their mean values follow the "Poisson
distribution" and SPSS has routines specifically designed for such
variables. In my experience it is common to have variances which
increase more rapidly than the means, a problem known in the trade as
"over-dispersion".

Some Poisson distributions (but certainly not all of them) can be fixed
by analysing the square root of the dependent variable.

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