Which of the following statements regarding seasonality is FALSE? A) Not correcting for seasonality when, in fact, seasonality exists in the time series results in a violation of an assumption of linear regression. B) If seasonality is quarterly, the appropriate term to add to the model is the previous year’s quarter’s time-series value. C) A time series that is first differenced can be adjusted for seasonality by incorporating the first-differenced value for the previous year’s corresponding period. D) The presence of seasonality makes it impossible to forecast using a time-series model.
Would it be C? If the seasonal quarter is Qt-4 the differencing would be (Qt-4 - Qt-5) ???
A? seasonality is not part of linear regression but part of the time series
Sorry misread the question. Gotta be A
a Linear regression assumes residuals are normally distributed, with E(e_t)=0. If there is seasonality, they won’t be. Given my recent record on quant, treat this answer with caution!
D
d
D is the answer. Well done maratikus. Can you please explain why? The answer given is Forecasting is no different in the case of seasonal component in the time-series model than any other forecasting.
Bloody READ THE QUESTION CHRIS!!! Which is FALSE! D. You can forecast with autoregressive model.
Time series has the potential to be a nasty vignette, imo.
I still feel the answer should be A. They were talking about violation of ‘linear regression’ and not ‘multiple regression’. NO???
Plus isn’t the point of adjusting for seasonality so that you can use it for forecasting?
Sponge_Bob_CFA Wrote: ------------------------------------------------------- > Plus isn’t the point of adjusting for seasonality > so that you can use it for forecasting? exactly! Without adjustment there will be serial correlation of residuals present -> violation of regression assumptions.
Hates life