ASYMPTOTICS AND CONFIDENCE ESTIMATION IN SEGMENTED REGRESSION MODELS

We consider confidence estimation methods for the change point parameter in a clustered segmented regression model with a continuity constraint using least squares estimators of the model parameters. The asymptotic properties of the estimators are first derived, and then these properties are used to demonstrate that the confidence estimation methods have the correct coverage probability. Finally, the performance of the methods is compared by simulation and the methods are applied to a real data example.