[R] Question

[Javier Bussi]

Dear Sir,

My name is Javier Bussi and I´m a professor at the University of Rosario,
Argentina. The National Household Survey has changed the periodicity of the
information provided on unemployment from biannual to quarterly. It is not
the usual situation where we have two series at the same period of time e.g
monthly and quarterly, and we combine both of them. In this case, we have a
biannual series that from a certain date changes to quarterly and we want to
combine them in order to make forecasts. Do you know any paper where this is
discussed? We could make the quarterly series become biannual, but we would
be loosing information, we think we can do better than that. Other approach
would be to apply structural models with continuos time, but we are trying
to find out if there are other alternatives.

Thank you very much for your time, 
Yours sincerely,

Javier Bussi

---