2. (15%) Using a data set including the weekly demand for hamburger, its price (x1) and whether it placed advertisement (x2) for the past two weeks, the marketing manager tried to fit a second-order linear model to predict weekly demand using the following regression model. Demand = Bo + B, Advertisement + B, Price + By (Price ) + E where Advertisement =0 when it did not place advertisement; and Advertisement = 1 when it placed advertisement The regression results are given below. SUMMARY OUTPUT Regression Statistics Multiple R 0.999078899 R Square 0.998158647 Adjusted R Square 0.997851754 Standard Error 13.9655485 Observations 22 ANOVA 55 MS F Significance F Regression 3 1903053.16 634351.0535 3252.47278 8.56846E-25 Residual 18 3510.657809 195.0365449 Total 21 1906563.818 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 4369.5 132.8897009 32.88065192 1.58301E-17 4090.309098 4648.690902 ADVERTISEMENT (X2) 70. 18181818 5.954929889 11.78549865 6.74307E-10 57.67097473 82.69266163 PRICE (X1) -146.6202214 6.759045534 -21.6924447 2.35925E-14 -160.8204492 -132.419994 PRICE SQ (X1) 1.26238345 0.084282879 14.97793453 1.32117E-11 1.085311691 1.439455209 DW = 1.9131) Is there any evidence to show the second-order model is useful in predicting Weekly Demand at 1% significance level? 2) Write down the corresponding second-order model with the estimated regression coefficients. 3) What is the effect of price on demand? 4) Explain the meaning of the coefficient of x2. 5) Perform the Durbin-Watson test for any possible serial correlation in the regression residuals