Competing risk models for turtle nest survival in the Bolivian Amazon
MetadataShow full item record
Survival analysis is a very common method to model time-to-event data. In most studies, the probability of experiencing one event (such as failure or death) is investigated. Sometime, multiple events (competing risk events) are of interest. Competing risk events can be estimated crudely by separately modeling single events, but more complete models are necessary to perform complete competing risk modeling. In this thesis, a competing risk model is used to analyze P.unifilis (yellow-spotted Amazon River Turtle) nest survival based on data collected from 1910 nests observed in the Bolivian Amazon in 2005 and 2006. Under this scenario, turtle nests experience risk from animals, floods and humans. The results from competing risk models are evaluated to show the risk event and risk period for turtle nests.