Y_t = 6.45 + 0.164Y_t-1 + 0.071Y_t-2 + 0.269Y_t-3

(1.79) (0.101) (0.068) (0.072)

where t is the annual number of worldwide earthquakes with a magnitude greater than 7 on the Richter scale in year . Standard errors are displayed in parenthesis and the time series is displayed below. The researcher is not quite sure whether he has included too many or too few lags in the model above. To deal with this problem, he tries AR(p) models with p ranging from 1 to 6 over the same sample period. The sum of squared residuals (SSR) from each of these estimated models is shown in the table below. Use BIC to estimate the number of lags that should be included in the regression. Do the results differ if you use AIC?

| | | 1 2 3 4 5 6 SSR 182.93 157.89 142.83 138.79 132.01 129.98 | | | 1 2 3 4 5 6 SSR 182.93 157.89 142.83 138.79 132.01 129.98