Some researchers3 illustrate a multiple linear regression equation to predict the yearly output of an oil field. We use their latest 20 years of data and change the units of three of their variables to
y = loge (yearly output) where output is measured in 100,000 barrels
x1 = loge (number of startup wells) where the number of new startup wells is measured in 1000.
x2 = loge (water volume) where last year's injected water volume is measured in 100,000 gallons
The least squared fit, to n = 20 years of data, gives the results

SS due to regression = 6.2696
SSE = .0431
(a) Predict the response for
(i) x1 = 1.5 and x2 = 2.5
(ii) x1 = 2.5 and x2 = 3.5
(b) Estimate the error standard deviation o and state the degrees of freedom.
(c) What proportion of the y variability is explained by the fitted regression?

A--.6415= .2869 =.6532