--- Jack Needleman wrote:
> I would have assumed that the correct model was a random effects logit,
> but in reviewing the literature came across two articles that argue that
> if you only have a single set of observations at the cluster level, the
> population averaged model is more appropriate:
>
> Neuhaus JM. Statistical methods for longitudinal and clustered designs
> with binary responses. Stat Methods Med Res. 1992;1(3):249-73. Review.
>
> Young ML, Preisser JS, Qaqish BF, Wolfson M. Comparison of
> subject-specific and population averaged models for count data from
> cluster-unit intervention trials. Stat Methods Med Res. 2007
> Apr;16(2):167-84.
>
> The results from the RE model are substantially closer to zero. The
> purpose of the research is to explore the association of one of the
> hospital level variables (staffing) on patient outcomes.
I found Chapter 13 of (Fitzmaurice et al. 2004) very helpful. Basicaly, their point is that these models have subtly different dependent
variables: the average probability of being discharged at the hospital
level (population average) versus the an individual's probability of
being discharged. So, choosing which model is correct for you depends on
who you want to describe (the individual or the hospital).
Hope this helps,
Maarten
Fitzmaurice, Garrett M., Laird, Nan M., Ware, James H. (2004) Applied Longitudinal Analysis. Hoboken, NJ: Wiley.
-----------------------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
