NERS 488/688 Homework 4
Date set 1.
- Use
the input file ners488hw3set1.inp in the Public folder, which contains capture
histories for a sample of animals.
The study was done over a
period of years when hunting regulations
were made progressively more liberal (longer seasons etc.). Managers are specifically interested in
whether changes in hunting regulations have affected adult survival.
There are four groups adult males and females and juvenile males and females.
- Use Program Mark to find the best (lowest AIC) model
for these data. Is there support
for an effect of hunting regulations?
Hint: You can build a model with a trend in survival by coding a
sequence of numbers in the column identifying a group. See Mark manual.
- Use Program Release to test the hypotheses of equal
catchability and equal survival probabilities for individuals in the
sample.
- Use the chi-square statistic/df from the total
chi-square from Tests 2 and 3 in Release to calculate c-hat. Use c-hat
to adjust AIC, producing QAIC scores.
- Produce parameter estimates for the best model.
- Produce model averaged estimates of parameters.
Data set 2.
- Use data set nres488tdhw3.inp from the Public folder. This data set
contains capture histories from about 100 individuals fro whom mass was
measured at the start of the study. Mass has been standardized to a
normal distribution with mean = 0 and SD = 1.
- Assess the role of mass in survival relative to other models for
survival. You may want to simplify models for capture probability
before considering models of survival. Hint: You will need to
incorporate mass as a covariate in the Design matrix to test effects of mass
on survival.
- Prepare a graph of the relationship between condition index and
survival. Hint: You will need to use the back-transformed logit and
the betas for the intercept and the covariate. Using the logit and the
range covariate calculate expected survival for a sample of covariate
values.