Mixed Model

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

Mixed Model

James Algina
Hello,

I am trying to help a student who is comfortable with SPSS to estimate a
mixed model that can be estimated in SAS.  I am not terribly  familiar
with SPSS and wonder if someone can help

The SAS code follows.  The variable score is a measure of attitude
towards deviance or  a measure of exposure to deviant peers. The
variable represented by score is specified by the dummy coded variables
attvar and expvar. Each participant has five attitude and five exposure
score. The intent of the model is to estimate the between person
variances for attitude and exposure, the between person covariance for
attitude and exposure and the within person variances for attitude and
exposure.


Proc mixed;
class varc id age11;
model score=attvar expvar /noint solution;
random     attvar expvar /subject=id type=un;
repeated /subject=id  group=varc;

The group=varc requests a separate within person variance for attitude
and exposure (at each level of varc).

This is what I have for SPSS syntax:

MIXED
   score  with attvar expvar
   /CRITERIA = CIN(95) MXITER(100) MXSTEP(5) SCORING(1) SINGULAR
  (0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE
  (0.000001, ABSOLUTE)
   /FIXED = attvar expvar| SSTYPE(3) noint
   /METHOD = REML
   /PRINT = G R SOLUTION
  /RANDOM attvar expvar| SUBJECT(id) COVTYPE(UN) .
/REPEATED|subject(id).


In SPSS, is there a way to modify the REPEATED command to a request
separate within person variance at each level of varc?

Thanks in advance.

Jamie

--
James Algina
Professor
Research and Evaluation Methodology
119D Norman Hall
P.O. Box 117047
University of Florida
Gainesville, FL 32611

Voice 352-273-4305
Fax   352-846-2697

=====================
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
Reply | Threaded
Open this post in threaded view
|

Re: Mixed Model

Ryan Black
James,

I remember seeing a Grouping keyword option when exploring the
GENLINMIXED procedure in SPSS 19, but that was quite some time ago. I
cannot say whether this keyword is equivalent to the GROUP= option in
SAS. Maybe, maybe not. Take a look at the documentation.

Ryan

On Fri, Mar 4, 2011 at 1:29 PM, James Algina <[hidden email]> wrote:

> Hello,
>
> I am trying to help a student who is comfortable with SPSS to estimate a
> mixed model that can be estimated in SAS.  I am not terribly  familiar
> with SPSS and wonder if someone can help
>
> The SAS code follows.  The variable score is a measure of attitude
> towards deviance or  a measure of exposure to deviant peers. The
> variable represented by score is specified by the dummy coded variables
> attvar and expvar. Each participant has five attitude and five exposure
> score. The intent of the model is to estimate the between person
> variances for attitude and exposure, the between person covariance for
> attitude and exposure and the within person variances for attitude and
> exposure.
>
>
> Proc mixed;
> class varc id age11;
> model score=attvar expvar /noint solution;
> random     attvar expvar /subject=id type=un;
> repeated /subject=id  group=varc;
>
> The group=varc requests a separate within person variance for attitude
> and exposure (at each level of varc).
>
> This is what I have for SPSS syntax:
>
> MIXED
>  score  with attvar expvar
>  /CRITERIA = CIN(95) MXITER(100) MXSTEP(5) SCORING(1) SINGULAR
>  (0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE
>  (0.000001, ABSOLUTE)
>  /FIXED = attvar expvar| SSTYPE(3) noint
>  /METHOD = REML
>  /PRINT = G R SOLUTION
>  /RANDOM attvar expvar| SUBJECT(id) COVTYPE(UN) .
> /REPEATED|subject(id).
>
>
> In SPSS, is there a way to modify the REPEATED command to a request
> separate within person variance at each level of varc?
>
> Thanks in advance.
>
> Jamie
>
> --
> James Algina
> Professor
> Research and Evaluation Methodology
> 119D Norman Hall
> P.O. Box 117047
> University of Florida
> Gainesville, FL 32611
>
> Voice 352-273-4305
> Fax   352-846-2697
>
> =====================
> 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