similar to: GEE - test for overdispersion and scale adjustment

Dear friends, As we all know, the usual model selection criteria(e.g.deviance,AIC...) in GLMs isn't very good for selecting the best model when overdispersion exist, so we need to adjust the corresponding statistic,see(Fitzmaurice,G.M. (1997) Model selection with overdispersed

Dear R users, I work with a descrete variable (sclae 0 - 27) which is highly skwed to the right (many zeros and small numbers). I measure this variable on a control and intervention cohort 5 times a year. When I analyze analyze this varoable at each time point separately and use GLM with family quasi-Poisson (descrete outcome and two binary variables, gender and cohort, are predictors), I observe

Hello! I am trying to model data on species abundance (count data) with a poisson error distribution. I have a fixed and a random variables and thus needs a mixed model. I strongly doubt that my model is overdispersed but I don't know how to get the overdispersion parameter in a mixed model. Maybe someone can help me on this point. Secondly, it seems that quasi models cannot be implemented

I have a few questions relating to overdispersion in a sex ratio data set that I am working with (note that I already have an analysis with GLMMs for fixed effects, this is just to estimate dispersion). The response variable is binomial because nestlings can only be male or female. I have samples of 1-5 nestlings from each nest (individuals within a nest are not independent, so the response

Hello, I am analysing counts data with a mixed model using lmer procedure. I therefore use the quasipoisson distribution but I'm not sure if this is sufficient to account for overdispersion. Actually the results are not very different to what I get when specifying a poisson distribution although my data are clearly overdispersed. this my model: >model <- lmer(NB ~ T + volume +

Hello, I am interested in running generalized estimating equation models in R. Currently there are two main packages for doing so in R, geepack and gee. I understand that even though one can obtain similar to almost identical results using either of the two, that there are differences between the packages. The paper that introduces the geepack package (

Hey Buddy, Hope you have been doing well since last contact. If you have the answer to the following question, please let me know. If you have chance to travel up north. let me know. best, -Sean ---------- Forwarded message ---------- From: Sean Zhang <seanecon@gmail.com> Date: Sat, Apr 11, 2009 at 12:12 PM Subject: question related to fitting overdispersion count data using lmer

Dear R-helpers: I have a question related to fitting overdispersed count data using lmer. Basically, I simulate an overdispsed data set by adding an observation-level normal random shock into exp(....+rnorm()). Then I fit a lmer quasipoisson model. The estimation results are very off (see model output of fit.lmer.over.quasi below). Can someone kindly explain to me what went wrong? Many thanks in

I am attempting to run a glm with a binomial model to analyze proportion data. I have been following Crawley's book closely and am wondering if there is an accepted standard for how much is too much overdispersion? (e.g. change in AIC has an accepted standard of 2). In the example, he fits several models, binomial and quasibinomial and then accepts the quasibinomial. The output for residual

Goal: use GEE or GLMM to analyze repeated measures data in R GEE problem: can?t find a way to do GEE with negative binomial family in R GLMM problem: not sure if I?m specifying random effect correctly Study question: Does the interaction of director and recipient group affect rates of a behavior? Data: Animals (n = 38) in one of 3 groups (life stages): B or C. Some individuals (~5)

Hi, having used both the gee pacakge and the geepack package, i am unsure of how to interpret the results. Here are the results from the geeglm function from the geepack package > gee2<-geeglm(data$erythema~data$product, data = data, id=subject, > family=binomial, corstr="independence") Warning message: In model.response(mf, "numeric") : using

Hi, there are a lot of messages dealing with overdispersion, but I couldn't find anything about how to test for overdispersion. I applied a GAM with binomial distribution on my presence/absence data, and would like to check for overdispersion. Does anyone know the command? Many thanks, Anna -- View this message in context:

Dear all, I need to fit a gee model with an auto-regressive correlation structure and I faced some problems. I attach a simple example: ####################################################### library(gee) library(geepack) # I SIMULATE DATA FROM POISSON DISTRIBUTION, 10 OBS FOR EACH OF 50 GROUPS set.seed(1) y <- rpois(500,50) x <- rnorm(500) id <- rep(1:50,each=10) # EXAMPLES FOR

I've been poking around with GLMs (on which I am *not* an expert) on behalf of a student, particularly binomial (standard logit link) nested models with overdispersion. I have one possible bug to report (but I'm not confident enough to be *sure* it's a bug); one comment on the general inconsistency that seems to afflict the various functions for dealing with overdispersion in GLMs

Hello Fellow R Users,I have spent the last week trying to find a work around to this problem and I can't seem to solve it. I simply want to plot my GEE model result with 95% confidence bands. I am using the geepack package to run a basic GEE model involving nestling weights, to a Gaussian distribution, with "exchangeable" error structure. I am examining how nestling weight varies

Hi, I am performing glm with binomial family and my data show slight overdispersion (HF<1.5). Nevertheless, in order to take into account for this heterogeneity though weak, I use F-test rather than Chi-square (Krackow & Tkadlec, 2001). But surprisingly, outputs of this two tests are exactly similar. What is the reason and how can I scale the output by overdispersion ?? Thank you,

Hi A quick question. Standard errors reported by gee/yags differs from the ones in geeglm (geepack). require(gee) require(geepack) require(yags) mm <- gee(breaks ~ tension, id=wool, data=warpbreaks, corstr="exchangeable") mm2 <- geeglm(breaks ~ tension, id=wool, data=warpbreaks, corstr="exchangeable", std.err = "san.se") mm3 <- yags(breaks ~

Hi, I have a bivariate longitudinal dataset. As an example say, i have the data frame with column names var1 var2 Unit time trt (trt represents the treatment) Now suppose I want to fit a joint model of the form for the *i* th unit var1jk = alpha1 + beta1*timejk + gamma1* trtjk + delta1* timejk:trtjk + error1jk var2 = alpha2 + beta2*timejk + gamma2* trtjk + delta2* timejk:trtjk +

Hi, I am trying to fit a ordinal GEE model using ordgee {geepack}. In order to check the validity of the function, I specified the correlation structure as independence (i.e. constr = "independence") and compared the result with that using polr {MASS}. Because a GEE model with an independent working correlation structure is equivalent to an ordinary GLM model, we would expect the same

Hi, I have to analyse the number of provisioning trips to nestlings according to a number of biological and environmental factors. I was thinking of building a mixed-effects model with species and nestid as random effects, using a Poisson distribution, but the data are overdispersed (variance/mean = 5). I then thought of using a mixed-effects model with negative binomial distribution, but I have