Determine differences. Use the DF test to determine a number of differences that will render the data stationary. 2. Generate Dickey-Fuller t-statistics manually. Check if the original series is stationary. Run the OLS equation that was estimated to calculate the DF (Dickey-Fuller) test statistic, and demonstrate that your estimated t-statistics value is the same as the DF value for 0 differences generated from part a). 3. Determine the optimal number of lags. Use SAC, PAC and SIC approaches to determine the optimal number of lags and compare the results. 4. Estimate the AR model. Use Simetar's "Time Series" wizard. Use the appropriate values using SAC from the previous step to specify the model. Ask Simetar for 5 "forecast periods