Complex samples - Subpopulation question

[John-220]

Hi all,
 I am new to SPSS complex samples and have a question regarding
subpopulation analysis...

My dataset of interest is a subpopulation of a large administrative
database. I have created dummy records that represent values from
the "missing" stara and clusters when I queried my set of interest. I have
an indicator variable (0=data of interest, 1= dummy records). I created a
csplan file on this dataset. When I run a regression using subpopulation
category = 0 the strata and cluster numbers printed are clearly only those
from the subpopulation (i.e. the strata and cluster numbers are less than
they should be). I am worried that the variances are not being calculated
with the correct degrees of freedom...Is this an issue or does SPSS
consider the entire design but only prints the values for the subpopulation?

Thanks in advance.

[Zdaniuk, Bozena]

Hello, I need to use MANOVA syntax to run a RM Ancova (mixed, one
btw-sbj, one within-sbj factor) with one covariate. I don't want to use
GLM because apparently it does not handle the covariates well in a mixed
btw, within design. Could someone point me to (or email me) an example
of MANOVA syntax for such design? :
1 btw subject factor A with two levels
1 within-subject factor B with two levels (time points)
1 continuous covariate.

Thanks a lot in advance.
Bozena

Bozena Zdaniuk, Ph.D.

University of Pittsburgh

UCSUR, 6th Fl.

121 University Place

Pittsburgh, PA 15260

Ph.: 412-624-5736

Fax: [phone]

email: [hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John
Sent: Tuesday, July 11, 2006 9:44 AM
To: [hidden email]
Subject: Complex samples - Subpopulation question

Hi all,
 I am new to SPSS complex samples and have a question regarding
subpopulation analysis...

My dataset of interest is a subpopulation of a large administrative
database. I have created dummy records that represent values from
the "missing" stara and clusters when I queried my set of interest. I
have
an indicator variable (0=data of interest, 1= dummy records). I created
a
csplan file on this dataset. When I run a regression using subpopulation
category = 0 the strata and cluster numbers printed are clearly only
those
from the subpopulation (i.e. the strata and cluster numbers are less
than
they should be). I am worried that the variances are not being
calculated
with the correct degrees of freedom...Is this an issue or does SPSS
consider the entire design but only prints the values for the
subpopulation?

Thanks in advance.

[Dale Glaser]

Bozena..here is a very brief example (note: time constant covariate is included in parenthesis......time varying covariate is not......i.e, Cov1 to Cov2:


  MANOVA DV1 DV2 by IV (1,2) with (Cov)
  /WSFACTORS=tm (2)
  /contrast (tm) = polynomial
  /print=transform Param(estim)
  /wsdesign=tm
  /design = IV.



  Dale

"Zdaniuk, Bozena" <[hidden email]> wrote:
  Hello, I need to use MANOVA syntax to run a RM Ancova (mixed, one
btw-sbj, one within-sbj factor) with one covariate. I don't want to use
GLM because apparently it does not handle the covariates well in a mixed
btw, within design. Could someone point me to (or email me) an example
of MANOVA syntax for such design? :
1 btw subject factor A with two levels
1 within-subject factor B with two levels (time points)
1 continuous covariate.

Thanks a lot in advance.
Bozena

Bozena Zdaniuk, Ph.D.

University of Pittsburgh

UCSUR, 6th Fl.

121 University Place

Pittsburgh, PA 15260

Ph.: 412-624-5736

Fax: [phone]

email: [hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John
Sent: Tuesday, July 11, 2006 9:44 AM
To: [hidden email]
Subject: Complex samples - Subpopulation question

Hi all,
I am new to SPSS complex samples and have a question regarding
subpopulation analysis...

My dataset of interest is a subpopulation of a large administrative
database. I have created dummy records that represent values from
the "missing" stara and clusters when I queried my set of interest. I
have
an indicator variable (0=data of interest, 1= dummy records). I created
a
csplan file on this dataset. When I run a regression using subpopulation
category = 0 the strata and cluster numbers printed are clearly only
those
from the subpopulation (i.e. the strata and cluster numbers are less
than
they should be). I am worried that the variances are not being
calculated
with the correct degrees of freedom...Is this an issue or does SPSS
consider the entire design but only prints the values for the
subpopulation?

Thanks in advance.



Dale Glaser, Ph.D.
Principal--Glaser Consulting
Lecturer--SDSU/USD/CSUSM/AIU
4003 Goldfinch St, Suite G
San Diego, CA 92103
phone: [phone]
fax: [phone]
email: [hidden email]
website: www.glaserconsult.com