Known fates input interval

[kd1221]

Hello all,

I am working on calculating survival for animals using GPS collar data. I will be using the Known-Fate method in Program Mark. Season is the shortest time interval for which we are interested in calculating survival. However, I am wondering, since the data is available, if there is any benefit to inputting the data in smaller intervals (e.g. monthly or daily).

Thank you very much for any help you can provide!

[cooch]

Hello all,

I am working on calculating survival for animals using GPS collar data. I will be using the Known-Fate method in Program Mark. Season is the shortest time interval for which we are interested in calculating survival. However, I am wondering, since the data is available, if there is any benefit to inputting the data in smaller intervals (e.g. monthly or daily).

Thank you very much for any help you can provide!

Depends on the time interval that is appropriate for the biology of the question. If you believe that 'interesting events' which influence survival occur at a weekly interval, then set the interval to a week. The main caveat is that as the interval gets shorter, period-specific survival -> 1, and you will have increasing difficulty (potentially) getting decent estimates (as you approach the 1.0 boundary). This is mitigated somewhat by the fact that with 'known-fate' data, you have a bit more wiggle room.