similar to: Robustness of linear mixed models

I'm fitting a GLMM to some questionnaire data. The structure is J individuals, nested within I areas, all of whom answer the same K (ordinal) questions. The model I'm using is based on so-called continuation ratios, so that it can be fitted using the lme4 package. The lmer function fits the model just fine, but using mcmcsamp to judge the variability of the parameter estimates produces

With lme4, use of mcmcsamp can be insightful. (Douglas Bates drew my attention to this function in a private exchange of emails.) The distributions of random effects are simulated on a log scale, where the distributions are much closer to symmetry than on the scale of the random effects themselves. As far as I can see, this is a straightforward use of MCMC to estimate model parameters; it is not

Hi! I am trying to calculate HPD for the coeficients of regression models fitted with lm or lmrob in R, pretty much in the same way that can be accomplished by the association of mcmcsamp and HPDinterval functions for multilevel models fitted with lmer. Can anyone point me in the right direction on which packages/how to implement this? Thanks for your time! R. [[alternative HTML version

This question is slightly off topic, but I'll use R to try and make it as relevant as possible. I'm working on a problem where I want to compare estimates from a posterior distribution with a uniform prior with those obtained from a frequentist approach. Under these conditions the estimates should agree. Specifically, I am asking the question, "What is the probability that the true

Hi folks, This is really minor but to someone not familiar with the various tentacles of the lmer package it could be really annoying. I was trying to plot the posterior density of the fixed effect parameters of a lmer model, > hr.mcmc = mcmcsamp(hr.lmer, n=50000) > densityplot(hr.mcmc, plot.points=F) There is this error, "Error in densityplot(hr.mcmc, plot.points = F) : no

From a quick look at the paper in the SAS proceedings, the simulations seem limited to nested designs. The major problems are with repeated measures designs where the error structure is not compound symmetric, which lme4 does not at present handle (unless I have missed something). Such imbalance as was investigated was not a serious issue, at least for the Kenward and Roger degree of freedom

Dear readers, Lm returns all information necessary to reconstruct summary statistics by group. However, t.test only returns the group means, and not the group SDs, or even the group Ns. These cannot be reconstructed from the test statistic and df, because the df are already pooled, except under a very strict assumption of equality of groups and variances. I need these summary statistics for a

Hello, I am a chemical student and I make use of 'lme/lmer function' to handle experiments in split-plot structures. I know about the mcmcsamp and I think that it's very promissory. I would like knowing "the concept behind" of the mcmcsamp function. I do not want the C code of the MCMCSAMP function. I would like to get the "pseudo-algorithm" to understanding that

I apologize if this question is not completely appropriate for this list. I have been using SAS for a while and am now in the process of learning some C and R as a part of my graduate studies. All of the statistical packages I have used generally yield p-values as a default output to standard procedures. This week I have been reading "Testing Precise Hypotheses" by J.O. Berger

Dear all, I am running a mixed model using lmer. In order to obtain CI of individual coefficients I use mcmcsamp. However, I need advice which values that are most appropriate to present in result section of a paper. I have not used mixed models and lmer so much before so my question is probably very naive. However, to avoid to much problems with journal editors and referees addicted to

I am changing the title because this is really about the history of anova, and about strata in analysis of variance. As this kind of question has been arising very frequently, an extended comment may be in order. The ideas, and the sums of squares breakdowns, go back to Fisher; see in particular his "Design of Experiments", first published in 1935. This book is still a good

Version 0.98-1 of the lme4 package and of the Matrix package are now on CRAN. This version provides the adaptive Gauss-Hermite quadrature (AGQ) method for fitting generalized linear mixed models. A new generic function mcmcsamp has been added with a method for objects in the "lmer" (linear mixed model fit or generalized linear mixed-effects model fit) class. This function provides a

Version 0.98-1 of the lme4 package and of the Matrix package are now on CRAN. This version provides the adaptive Gauss-Hermite quadrature (AGQ) method for fitting generalized linear mixed models. A new generic function mcmcsamp has been added with a method for objects in the "lmer" (linear mixed model fit or generalized linear mixed-effects model fit) class. This function provides a

I have written a function called mcsamp() that is a wrapper that runs mcmcsamp() and automatically monitors convergence and structures the inferences into vectors and arrays as appropriate. But I have run into a very little problem, which is that mcmcsamp() shortens the variable names. For example: > set.seed (1) > group <- rep (1:5,10) > a <- rnorm (5,-3,3) > y <-

OK, I see now that I was supposed to twig that the reference was to putting the ?.Rnw' files back into the vignettes directory from the inst/doc directory where they?d been placed in the course of creating the tar.gz file. I am still trying to work out what I need to put into ?.Rinstignore? so that ?.install_extras? is not installed. John Maindonald email: john.maindonald at

I am helping my wife do some statistical analysis. She is a biologist, and she has performed some measurements on various genotypes of mice. My background is in applied mathematics and engineering, and I have a fairly good statistics background, but I am by no means a PhD level expert in statistical methods. We have used the lmer package to fit various models for the various experiments that she

Sorry. This, and the description in the ?Writing R Extensions? manual, leaves me completely mystified. Is it that I have to remove the PDFs that are created when I run ?R CMD build?, and somehow ensure that they are rebuilt when the package is installed? Do I need a Makefile? John Maindonald email: john.maindonald at anu.edu.au<mailto:john.maindonald at anu.edu.au> phone :

Hello, I built 4 mixed models using different data sets and standardized variables as predictors. In all the models each of the fixed effects has an associated random effect (same predictor). What I find is that fixed effects with larger (absolute) standardized parameter estimates have also a higher estimate of the related random effect. In other words, the higher the average of the absolute

You've omitted a comma. races2000 is a data frame, which for purposes of extracting rows behaves like a 2-dimenional object. The following works fine: hills2000 <- races2000[races2000$type == 'hill', ] Additionally, you might like to ponder > type <- races2000[names(races2000)=="type"] > type[1:4] Error in "[.data.frame"(type, 1:4) :

Hello, R community, I have been using the lmer and mcmcsamp functions in R with some difficulty. I do not believe this is my code or data, however, because my attempts to use the sample code and 'sleepstudy' data provided with the lme4 packaged (and used on several R-Wiki pages) do not return the same results as those indicated in the help pages. For instance: > sessionInfo() R