Ljung-Box Q-statistic
Message posted by Tomi Owens (via 154.32.143.114) on August 21, 2001 at 12:42 PM (ET)

I know that the Q-statistic has a chi-square distribution with k-m degrees of freedom where k is the number of lags and m is the number of estimated parameters. I have a time series that I am modelling by classical seasonal decomposition. I have 8 years of 48 weekly observations (48 week school year). I therefore have seasonality with period 48.

The big question - have I estimated 48 parameters (the seasonal components) or have I estimated 2 parameters (the trend-cyclical component and the seasonal component)?

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: Ljung-Box Q-statistic
Message posted by farahrasheed (via 203.135.0.66) on December 7, 2001 at 12:28 PM (ET)