www.cmm.bristol.ac.uk/forum

Thanks for the link, Chris. Interesting article. As the article states, we have theoretical reasons to test the statistical efficiency of a multilevel zero-inflated negative binomial model and a regular multilevel negative binomial model. I assume you or Kelvyn are not aware of any (R) package that ...

Hello,

Can Mlwin fit zero-inflated negative binomial models? I know that we can fit Poisson and negative binomial models.

thanks in advance
Adel

Thanks, again, Chris. Regarding the behavior of the predict() function. 1) does it take the sample data stored in the mymodel object? 2) not getting what the difference between the terms= argument and the params= argument (from the manual and my own experimentation)? Could you please help me underst...

Thanks for the clarifications, Chris, and some additional questions: 1) Does the predict(mymodel) function takes the mean of the covariates or the actual data values for each observation (and plug them into the model)? [it looks like it is taking the actual values…] What if I want to set some covari...

ok, strange, because I specified resi.store=T, with the following options:

resi.store = TRUE, resioptions= c("standardised", "leverage", "influence", "deletion"),

Should I specify the "resioptions" differently?

Thanks George for the reply. So the “the standardised residual x fixed part prediction” in the Mlwin enviroment should show me this, right?

Best,
Adel

Hi I would like to check the homoscedasticity assumption in my logistic multilevel model. According to this post https://www.cmm.bris.ac.uk/forum/viewtopic.php?f=1&t=2006&e=1&view=unread#p4567 , “You can assess the homoskedasticity assumption by plotting the predicted random effects agai...

Hi Chris, I am not sure if this is the code that is causing an error. The following is my model, with the residuals stored: > myresi <- result.l_imf.educ.collapsed.urban.int[[1]]@residual > names(myresi) [1] "lev_2_resi_est_Intercept" "lev_2_resi_se_Intercept" "lev_2_std_res...

Hi 1) It is usually suggested that it is not meaningful to check the homoscedasticity assumption in single level logistic models because the outcome is binary, but is it meaningful to check homoscedasticity on the log-odds scale? Probably the answer is still no, since one does not assume homoscedast...

Hi I would like to draw Quantile-Quantile plots to check model assumptions. But when--and this is not uncommon-- I have missing observations, the caterpillarR throws an error at me: > caterpillarR(mymodel['residual'], lev = 2) Error in tt[1, 1, ] <- var : number of items to replace is not a multiple...