Dear List,
I have two samples and have fitted an AR(1) model to both of them: no constant or time trend. Both series are stationary. Can anyone please provide me with a reference to the literature or with an explanation of a test of the hypothesis of whether or not rho (the autoregressive parameter) is the same in both samples?
Does the test change if I am using an AR(1) with a constant in one sample, and a zero mean AR(1) in another?
I am very thankful for this.
Carlos Santos
Clix Rapidix - Aumente até 6X a velocidade da sua Internet
Adira em http://acesso.clix.pt e comece logo a navegar
|