6/3/2023 0 Comments Brother pe design 10 rar![]() If the sampling weights are ignored at either level the parameter estimates can be substantially biased. Sampling weights are assigned at one or both levels to reflect these probabilities. Such sampling designs however often use unequal probability of selection at the cluster level and at the individual level. Multilevel models are frequently used to analyze data from cluster sampling designs. The lme4 package is great for multilevel modeling, but there is not a way that I know to include weights at different levels of clustering. I've used the survey package to weight for unequal probabilities of selection in one-level models, but this package does not have functions for multilevel modeling. ![]() I'm looking for advice on how to analyze complex survey data with multilevel models in R.
0 Comments
Leave a Reply. |