Logistic sensitivity and bounds for extinction risks

A method of sensitivity analysis for stochastic population models is proposed in which logistic regression is used to relate the probability of population decline to the parameters of the model. The regression equation serves to summarise the effects of different model parameters and interactions. The method was applied to a model of the helmeted honeyeater. The results may be used to place bounds on the predictions of the model. The predicted risk of population decline within the next 50 years is subject to substantial error.