Mixed models, Repeated Measures, multiple DVs, SPSS

[lir6777]

Hi there,

 I'm looking for some advice on mixed models please? My data is approximately as follows:

 400 participants split over 2 different attitude interventions. The participants are school students from 14 different schools so we want to use mixed models to allow for this nested data.

 The pp's were measured on 4 attitude items (which will be somewhat correlated but are measuring quite different things)both before and after the intervention, so I have 4 continuous DVs and repeated measures. I want to compare the effect the two interventions had on attitudes. I essentially want to use the pre-intervention measures as covariates.

However, in SPSS, the mixed models- linear model- doesn't seem to let you input more than one DV- anyone know what I should do here? I'm reluctant to meld the 4 DVs into one composite one as the measures are quite different and I would lose a lot of good information here..

 Any advice appreciated! Thanks!

[Bruce Weaver]

Here's an example with 2 DVs and a single explanatory variable. You should be able to do something similar for your situation.

http://spssx-discussion.1045642.n5.nabble.com/How-to-investigate-the-association-between-two-series-of-measurements-tp5720365p5720373.html

HTH.

lir6777 wrote

Hi there,

 I'm looking for some advice on mixed models please? My data is approximately as follows:

 400 participants split over 2 different attitude interventions. The participants are school students from 14 different schools so we want to use mixed models to allow for this nested data.

 The pp's were measured on 4 attitude items (which will be somewhat correlated but are measuring quite different things)both before and after the intervention, so I have 4 continuous DVs and repeated measures. I want to compare the effect the two interventions had on attitudes. I essentially want to use the pre-intervention measures as covariates.

However, in SPSS, the mixed models- linear model- doesn't seem to let you input more than one DV- anyone know what I should do here? I'm reluctant to meld the 4 DVs into one composite one as the measures are quite different and I would lose a lot of good information here..

 Any advice appreciated! Thanks!

[Ryan]

After restructuring the dataset into vertical format such that multiple response variables are collapsed into a single response vector, the following MIXED code can be employed to model the multivariate response, taking into account the specifications provided in the original post:

mixed y by y_indic condition with x
/fixed = y_indic y_indic*condition y_indic*x | noint sstype(3)
/method = reml
/print = g r solution
/random = y_indic | subject(school_id) covtype(un)
/repeated = y_indic | subject(student_id) covtype(un).

 

where

 

y = response vector (response-specific post scores)

y_indic = response indicator

condition = condition indicator

x = response-specific pre scores

school_id = school identifier

student_id = student identifier

 

I could spend a great deal of time describing this multivariate model and its assumptions (e.g., MVN distribution), along with specific tests that may be performed via the TEST sub-command (e.g. multivariate response effect of condition, response-specific effects of condition). Instead, I will leave it to the OP to explore the code and write back with any questions.

Ryan

On Fri, Jun 7, 2013 at 6:31 AM, Bruce Weaver <[hidden email]> wrote:

Here's an example with 2 DVs and a single explanatory variable. You should be
able to do something similar for your situation.

http://spssx-discussion.1045642.n5.nabble.com/How-to-investigate-the-association-between-two-series-of-measurements-tp5720365p5720373.html

HTH.


lir6777 wrote

> Hi there,
>
>  I'm looking for some advice on mixed models please? My data is
> approximately as follows:
>
>  400 participants split over 2 different attitude interventions. The
> participants are school students from 14 different schools so we want to
> use mixed models to allow for this nested data.
>
>  The pp's were measured on 4 attitude items (which will be somewhat
> correlated but are measuring quite different things)both before and after
> the intervention, so I have 4 continuous DVs and repeated measures. I want
> to compare the effect the two interventions had on attitudes. I
> essentially want to use the pre-intervention measures as covariates.
>
> However, in SPSS, the mixed models- linear model- doesn't seem to let you
> input more than one DV- anyone know what I should do here? I'm reluctant
> to meld the 4 DVs into one composite one as the measures are quite
> different and I would lose a lot of good information here..
>
>  Any advice appreciated! Thanks!

-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Mixed-models-Repeated-Measures-multiple-DVs-SPSS-tp5720624p5720625.html

Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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