In Exercise 30 we looked at the weekly average retail price (cents per gallon) of regular gas nationwide from 2011 through June 2013. Here’s the time series plot again:

a) What components can you see in this plot?
Here’s an AR(14) model fit to these data.
Dependent variable is: Gas Price 130 total cases of which 14 are missing R squared = 31.0% R squared (adjusted) = 30.4%
s = 0.1587 with 116 - 2 = 114 degrees of freedom Variable Coefficient SE(Coeff) t-ratio P-value Intercept 5.34842 0.2354 22.7 60.0001 Lag14 -0.463680 0.0647 -7.16 60.0001 Here’s a time series plot of the model.

b) Does this model show that there is a (possibly unsuspected)
14-week seasonal cycle in gas prices? Explain.

c) Would you use this model to predict future gas prices?
Explain.