I.  Objective:  Replicate Example  9.3 in Studenmnund text on pp. 315-324 analyzing Serial Correlation, the Durbin-Watson Test, and corrections for Serial Correlation using the chick6.xls data file discussed in the text.   The chick6.xls file contains US annual information (from 1974-2002) on the following variables:  Y (per capita chicken consumption in pounds), PC (price of chicken in cents/lb.), PB (price of beef in cents/lb.) and YD (per capital income in 100s of $). Assignment: 1.   Generate descriptive statistics for the Y, PC, PB and YD variables using EALimdep (click here for help).   What do they tell us about the average values of the variables and how much they vary? 2.    Run an ordinary least squares (OLS) regression of Y on PC, PB and YD, and plot the error terms over the differing time periods (after you specify the Y, ONE, PC, PB and YD variables in the regression, click on the OUTPUT tab and click on <Plot residuals><Against variable>< ONE>).    Does the plot suggest serially correlated errors? 3.   Write an equation that describes the regression model, assuming that the errors are serially correlated.  Test whether the ordinary least squares regression results suffers from serial correlation using the Durbin-Watson test. 4.    If the error terms are serially correlated, what would be the likely consequences for the estimated OLS regression results? 5.    Reestimate the regression, using the Cochrane-Orcutt correction for serial correlation (after you specify the Y, ONE, PC, PB and YD variables in the regression, click on the OUTPUT tab and click on the <OPTIONS> tab, and then click on <Autocorrelation><Correct for Autocorrelation using><Cochrance Orcutt>.  Compare the corrected results to the OLS results. 6.  Reestimate the regression, using the Prais-Winsten correction for serial correlation (after you specify the Y, ONE, PC, PB and YD variables in the regression, click on the OUTPUT tab and click on the <OPTIONS> tab, and then click on <Autocorrelation><Correct for Autocorrelation using><Prais-Winsten>.  Compare the corrected results to the OLS results and the Corchrane-Orcutt results. 7.   EALimdep doesn’t print out an R squared for the corrected results but you can compute the R squared.  Note that the R squared = 1 - (Unexplained Variation / Total Variation).  The Total Variation can be found by squaring the standard deviation of the dependent variable (Y) and then multiplying by N-1.  The Unexplained Variation can be found by squaring the standard deviation of residuals and then multiplying by N-#coefficients.  Compute the R squareds for both the OLS, the Cochrane-Orcutt and the Prais-Winsten results and compare. II.  Objective: