Simulating multilevel dataset

4 messages · Stuart Luppescu, Thompson,Paul

Original

1

4

Stuart Luppescu

Fri, Jul 5, 2013 10:18 AM #

On Fri, 2013-07-05 at 16:53 +0000, Thompson,Paul wrote:

The model would look something like this (in lmer notation): meas ~ var1.indic + var2.indic + (var1.indic|school) + (var2.indic|school)
Does this address your questions?

_______________________________________________
R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

-----------------------------------------------------------------------
Confidentiality Notice: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and may contain
privileged and confidential information.  Any unauthorized review, use,
disclosure or distribution is prohibited.  If you are not the intended
recipient, please contact the sender by reply e-mail and destroy
all copies of the original message.

Stuart Luppescu <slu at ccsr.uchicago.edu>
University of Chicago

Fri, Jul 5, 2013 10:25 AM #

Well, that's how it would be analyzed. But you have means and variability measures to generate the data. Do you have any notion of the dependencies/conditionalities inherent in the data? Are conditionalities additive? Are you considering some variance effects?

Prior to modeling conditionality, you need parameters to model in the direction of.

-----Original Message-----
From: Stuart Luppescu [mailto:slu at ccsr.uchicago.edu] 
Sent: Friday, July 05, 2013 12:19 PM
To: Thompson,Paul
Cc: r-sig-mixed-models
Subject: RE: [R-sig-ME] Simulating multilevel dataset

On Fri, 2013-07-05 at 16:53 +0000, Thompson,Paul wrote:

The model would look something like this (in lmer notation): meas ~ var1.indic + var2.indic + (var1.indic|school) + (var2.indic|school) Does this address your questions?

_______________________________________________
R-sig-mixed-models at r-project.org mailing list 
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

----------------------------------------------------------------------
- Confidentiality Notice: This e-mail message, including any 
attachments, is for the sole use of the intended recipient(s) and may 
contain privileged and confidential information.  Any unauthorized 
review, use, disclosure or distribution is prohibited.  If you are not 
the intended recipient, please contact the sender by reply e-mail and 
destroy all copies of the original message.

--
Stuart Luppescu <slu at ccsr.uchicago.edu>
University of Chicago


-----------------------------------------------------------------------
Confidentiality Notice: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and may contain
privileged and confidential information.  Any unauthorized review, use,
disclosure or distribution is prohibited.  If you are not the intended
recipient, please contact the sender by reply e-mail and destroy
all copies of the original message.

Stuart Luppescu

Fri, Jul 5, 2013 10:25 AM #

On Fri, 2013-07-05 at 12:18 -0500, Stuart Luppescu wrote:

Oops. I forgot to indicate that there is no intercept. I guess I should
have put a -1 in there.

Stuart Luppescu -=-=- slu <AT> ccsr <DOT> uchicago <DOT> edu
CCSR at U of C ,.;-*^*-;.,  ccsr.uchicago.edu
     (^_^)/    ????????
[Crash programs] fail because they are based on the theory that, 
with nine women pregnant, you can get a baby a month.
                -- Wernher von Braun

Stuart Luppescu

Fri, Jul 5, 2013 10:48 AM #

On Fri, 2013-07-05 at 17:25 +0000, Thompson,Paul wrote:

The variance-covariances at the individual level are
 3.308523 1.896117
 1.896117 5.694732

at the school level they are

      0.75137       0.59118 
      0.59118       0.93246

Not sure what you mean here. I'm not assuming there are any cross-level
interactions, if that's what you're referring to.

Again, I apologize for my ignorance, but I don't understand this. The
only parameters in the model are the individual means, variances and
covariances; and the school means, variances and covariances. Do I need
something else?

Stuart Luppescu <slu at ccsr.uchicago.edu>
University of Chicago