Open ycanbolat opened 8 months ago
Thanks for the request @ycanbolat it seems the formula for the unweighted case is 59 in the lme4 vignette. vignette(package="lme4", "lmer")
or in the raw file here. Though I haven't thought about how weighting would change that.
Looking at this again, I think the weighted version is the same as the unweighted version. That would make this relatively easy to implement, and that conclusion can be tested with our existing tests.
I`m using WeMix to analyze the PISA data. I fit a three-level model and want to extract random effects and associated standard errors by units (e.g., random effects and standard errors country by country for instance ). I was able to extract the random effects from "ranefMat" but could not find the associated standard errors. It will be great if the WeMix includes this function.