Figure 4.
Results from a mixed model with 10 regressors (otherwise known as fixed effects), the first five of which determine the observed data, and one random effect for each individual. (a) Model selection measures based on models with a varying number of regressors Jsel, where only those regressors Jsel and below are included (hence, Jsel = 5 represents the true model). Here, R = 200. (b) The model selection measures relative to the true model (for clarity, a table showing these values has been placed into electronic supplementary material, appendix O). (c) CPU time necessary to accurately estimate the model selection measures (within an uncertainty of 0.2) as a function of the number of observations R.