Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Assistance with xtmixed - categorical predictors, unequal observation intervals


From   "Benigno Rodriguez G., MD" <rodriguez.benigno@clevelandactu.org>
To   "statalist" <statalist@hsphsun2.harvard.edu>
Subject   st: Assistance with xtmixed - categorical predictors, unequal observation intervals
Date   Wed, 29 Oct 2008 16:59:10 -0400

I am reposting this with a modified subject line, as it did not show up on the list, and it originally contained "Help" as the first word of the subject, which is apparently problematic. I apologize if this ends up duplicated!

Hi,

I have a dataset that contains CD4 measurements on 400 unique persons at 10 unevenly spaced timepoints (2,4,16,40 weeks et cetera). I am interested in fitting a random intercept and slope model to these data, using the fixed effect "region" (categorical with 4 levels) and the random effect "BaselineCD4". The variable "week" contains the time, and the dataset is in the "tall" format. Could I please request help with the proper syntax?

The questions I want to address are:

1.- Does the estimated slope of CD4 vary by region (controlling for baselineCD4)? If so, which specific regions have an estimated larger of smaller slope (compared with region 1) 2.- Can I obtain an estimate of the magnitude of the effect of region, comparing each region to the reference (region=1)? 3.- Which covariance structure is best suited to the data? The correlation matrix is shown below. 4. The initial graphic assessment suggests that a quadratic model could fit the data better (an initial rapid increase in CD4, followed by a slower slope. How would this model be fitted?

. corr c4wk0 c4wk2 c4wk4 c4wk8 c4wk12 c4wk16 c4wk24 c4wk32 c4wk40 c4wk48
(obs=296)

| c4wk0 c4wk2 c4wk4 c4wk8 c4wk12 c4wk16 c4wk24 c4wk32 c4wk40 c4wk48
-------------+------------------------------------------------------------------------------------------
       c4wk0 |   1.0000
       c4wk2 |   0.8439   1.0000
       c4wk4 |   0.8400   0.8073   1.0000
       c4wk8 |   0.8112   0.7860   0.7757   1.0000
      c4wk12 |   0.8325   0.7969   0.8045   0.8269   1.0000
      c4wk16 |   0.8391   0.7715   0.7604   0.7960   0.8663   1.0000
      c4wk24 |   0.7731   0.7432   0.7551   0.7578   0.8095   0.7855   1.0000
c4wk32 | 0.7899 0.7499 0.7753 0.7581 0.8289 0.7894 0.8544 1.0000 c4wk40 | 0.7915 0.7430 0.7659 0.7717 0.8131 0.8075 0.7934 0.8545 1.0000 c4wk48 | 0.7186 0.6755 0.6958 0.7451 0.7765 0.7487 0.7697 0.8338 0.8612 1.0000

Any help would be tremendously appreciated!

Best,

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index