Regression Analysis of Time Series Data
= 24.6 − 0.0892x
. The residual plot versus time indicates there is autocorrelation.
b. d = .81 which rejects the null hypothesis and indicates that there is evidence of positive autocorrelation.
c. We get
The new regression equation is
′ = 12.0854 − 0.1105x
The standard errors of the regression coefficients are se
) = 0.5542 and se
) = 0.0.01403.
d. d = .90 which indicates there is still evidence of positive autocorrelation.
The regression through the origin for the first difference approach yields an estimataed slope of 0.28943 with a standard error of 0.02508. The previous estimate for 1
was 0.29799 with a standard error of 0.0123. As a result, the estimates are very similar, but the standard error is smaller for the ...