Wl. Kendall et Je. Hines, Program RDSURVIV: an estimation tool for capture-recapture data collected under Pollock's robust design, BIRD STUDY, 46, 1999, pp. 32-38
Several papers have demonstrated the advantages of collecting capture-recap
ture data using subsamples (i.e. Pollack's robust design). Compared with a
standard design (i.e. one sample per period), this approach (1) permits the
estimation of more demographic parameters and (2) in many cases produces m
ore efficient estimators. Program SURVIV is a powerful tool for computing p
arameter estimates under the robust design. However multinomial models deve
loped for the robust design require cells for each possible capture history
that occurs across the subsamples within a period. This makes the process
of entering cell probabilities in SURVIV very tedious and subject to errors
. Program RDSURVIV combines SURVIV with a front end that converts capture h
istories taken under the robust design to the proper input format for estim
ating parameters under a general model, nlm builds that model. This model p
ermits Markovian temporary emigration, trap response in capture probabiliti
es and time variation in all parameters. Program RDSURVIV also automaticall
y computes estimates under a series of submodels, but also permits the user
to specify other submodels.