Table 1 Summary of our simulation study results (in percentages) for three design questions focusing on ability of the Location-based Change Point Model (LCPM) to correctly detect the timing of the change for and the magnitude of change in geographic locations (magnitude), duration of observation following the event occurring (duration), and fix interval (frequency)
From: Change-point models for identifying behavioral transitions in wild animals
Category | Magnitude | Duration | Frequency | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
100% Difference | 75% Difference | 50% Difference | 25% Difference | 48 H | 24Â h | 12Â h | 06Â h | 03Â h | 15Â min | 30Â min | 60Â min | |
Level 1 | 100 | 98 | 94 | 14 | 100 | 100 | 100 | 100 | 100 | 100 | 8 | – |
Level 2 | – | 2 | – | 24 | – | – | – | – | – | – | 92 | – |
Level 3 | – | – | 6 | 4 | – | – | – | – | – | – | – | – |
Not successful | – | – | – | 56 | – | – | – | – | – | – | – | – |
No change detected | – | – | – | 2 | – | – | – | – | – | – | – | 100 |
Did not converge | – | – | – | – | – | – | – | – | – | – | – | – |