Dear All, I'm analysing a negative binomial dataset from a population-based study. Many covariates were determined on household level, so all members of a household have the same value for those covariates. In STATA, there seems to be an option for 'clustered analysis' for neg-bin regression. Does an equivalent exist for R(MASS)'s glm.nb or a comparable function? Many thanks for all help! Lutz -- Lutz Ph. Breitling, Marie Curie EST Research Fellow Integrative & Comparative Biology (Irene Manton 8.93e) University of Leeds, LS2 9JT Leeds/UK
I'm not certain, but I think you should be able to use the robcov function in the Design package in conjunction with glm.nb, perhaps with a bit of tweaking. Another alternative might be to use the quasi family in lmer. On 06/11/06, Lutz Ph. Breitling <lutz.breitling at gmail.com> wrote:> Dear All, > > I'm analysing a negative binomial dataset from a population-based > study. Many covariates were determined on household level, so all > members of a household have the same value for those covariates. > > In STATA, there seems to be an option for 'clustered analysis' for > neg-bin regression. Does an equivalent exist for R(MASS)'s glm.nb or a > comparable function? > > Many thanks for all help! > Lutz > -- > Lutz Ph. Breitling, Marie Curie EST Research Fellow > Integrative & Comparative Biology (Irene Manton 8.93e) > University of Leeds, LS2 9JT Leeds/UK > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >-- ================================David Barron Said Business School University of Oxford Park End Street Oxford OX1 1HP
Lutz Ph. Breitling said the following on 2006-11-06 13:56:> Dear All, > > I'm analysing a negative binomial dataset from a population-based > study. Many covariates were determined on household level, so all > members of a household have the same value for those covariates. > > In STATA, there seems to be an option for 'clustered analysis' for > neg-bin regression. Does an equivalent exist for R(MASS)'s glm.nb or a > comparable function?Which Stata function are you referring to? If you want a Poisson regression with gamma distributed random effects in R, I think you're out of luck. HTH, Henric> > Many thanks for all help! > Lutz
The R-package glmmADMB (not on CRAN) can do negative binomial response with one level of Gaussian random effects. It also allows zero inflation. Download site: http://otter-rsch.com/admbre/examples/glmmadmb/glmmADMB.html Regards, hans>Dear All, > >I'm analysing a negative binomial dataset from a population-based >study. Many covariates were determined on household level, so all >members of a household have the same value for those covariates. > >In STATA, there seems to be an option for 'clustered analysis' for >neg-bin regression. Does an equivalent exist for R(MASS)'s glm.nb or a >comparable function? > >Many thanks for all help! >Lutz[[alternative HTML version deleted]]
Lutz-- In addition to the previous suggestions, you might also have a look at the following paper: Kelvin K. W. Yau, Kui Wang and Andy H. Lee. Zero-Inflated Negative Binomial Mixed Regression Modeling of Over-Dispersed Count Data with Extra Zeros. Biometrical Journal 45 (2003) 4, 437?452 The authors have written R/Splus code (available from them). Although the article is focused on ZINB models with random-effects, the code will fit a NegBin with random-effects (if memory serves). Hope that helps. cheers, Dave -- Dave Atkins, PhD Assistant Professor in Clinical Psychology Fuller Graduate School of Psychology Email: datkins at fuller.edu Phone: 626.584.5554