odprism: Optimal design and performance of random intercept and slope models

Simulation functions to assess the optimal design and performance of random intercept and slope models (i.e. mixed models), which can be used to a priori determine adequate sampling designs for e.g. reaction norm studies. Functions allow users to vary the sampling design in terms of number of grouping units sampled (e.g. individuals, schools, populations, etc.) and replicates per grouping unit (unbalanced as well balanced datasets) and also allow users to vary the parameter conditions used to generate the data. Subsequently, the performance of mixed models (based on lme4 package) fitted on these datasets is assessed in terms of the accuracy and the precision of estimates of fixed and random parameter, as well as the statistical power.

Downloads: