Table 4. Bayesian Model Selection Results for Daily Sampling Interval Models Fit to Simulated Data.
| Model | Quarterly | Monthly | Weekly | Daily | ||||
|---|---|---|---|---|---|---|---|---|
| Top Model % | Pr(Model|data) | Top Model % | Pr(Model|data) | Top Model % | Pr(Model|data) | Top Model % | Pr(Model|data) | |
| Distance | 0 | 0 (0,0) | 0 | 0 (0,0) | 0 | 0 (0,0) | 0 | 0 (0,0) |
| Sex | 71.9 | 0.996 (0, 1.000) | 66.3 | 0.998 (0, 1.000) | 60.7 | 0.944 (0, 1.000) | 64.0 | 0.997 (0, 1.000) |
| σsex | 2.2 | 0 (0, 0.2778) | 0 | 0 (0,0) | 0 | 0 (0,0) | 0 | 0 (0,0) |
| Sex + σsex | 25.8 | 0.002 (0, 1.000) | 33.7 | 0.002 (0, 1.000) | 39.3 | 0.056 (0, 1.000) | 36.0 | 0.003 (0, 1.000) |
Summary of Bayesian model selection results for all simulated datasets with daily, weekly, monthly and quarterly sampling intervals, and Distance, Sex, σsex and Sex + σsex models. Top Model % gives the percent of simulations that identified a particular model as the having the highest posterior probability. Pr(Model|data) gives the median posterior probability estimate of a model given the data within the model set with 95% credibility interval of this probability in parentheses. The Sex model represents the true data generating process for all simulations.