Effective sample size in a data set

[Giacomo]

Hi Remi,
just a quick question to check my results.
When correcting the AIC for the sample size (the AICc), how does ESURGE calculate the effective sample size from a data set ?
Thank you very much
Giacomo

[CHOQUET]

Hello Giacomo,
The effective sample size used in E-SURGE is the total number of capture minus the number of individual retired.
Sincerely,
Rémi

[Giacomo]

Remi,

thank you very much.
I have a sample size of 444 observations (including the first one) of 143 individuals (none is censured or retired).
ESURGE considers an effective sample size of 444 while MARK of 393. Model selection is not really affected by this, but considering it is a small sample size, I was puzzled. Unfortnately I was not able to find much information on this topic.
Any idea why this difference ?
Do you (or anyone at this Forum) have any reference to suggest on how to calculate the effective sample size in CR dataset ?

Anyway, thank you very much for the answer
Giacomo

[Morten Frederiksen]

HI Giacomo,

ESURGE considers an effective sample size of 444 while MARK of 393.

In MARK, the effective sample size is calculated as the number of releases, so it doesn't include observations at the last occasion. If you had 51 observations at the last occasions, this adds up.

Which of these estimates of effective sample size (444 or 393) is more correct I will leave up to the more theoretically minded folks ...

Morten