What is the difference between the 2 models know that the seasonality model eradicates the autocorrelation due to seasonality by adding a lag. But the AR2 also does the same. The only difference is that seasonality uses log on the formula and AR2 doesn’t, right?
Either one can use logs; it depends on the data.
If there are only 2 seasons, then the models are identical. If there are more than 2 seasons, then the seasonality model will have a lag of more than 2 periods, while an AR2 model will not.
@RickSanchez and @S2000magician: Wouldn’t the difference between AR2 and Seasonality be the period of the lag such that:
a) the additional lag in the AR2 needs to be the period right before the lag in the AR1 model (i.e. adding t-2); and
b) the lag in the Seasonality model the period 1 year before the dependent variable (i.e. adding t-4 in the case of the quarterly data)
Agree that the use of logs is not a difference between the two models.
AR model is used to check the presence of Serial correlation in a time series. Whereas seasonality model is to check if there is any serial correlation in a particular year or quarter. The distinguishing point between the two models is, seasonality model uses lagged variables upto a year, unlike AR model