Table 3: Durbin Watson test on the autocorrelations in the PSCL model residuals.

SeriesDWLevel resultResult

Year2.104-D(U) > DW > D(U)No autocorrelation
Winter2.444-D(U) > DW > D(U)No autocorrelation
Spring0.56DW < D(L)Positive autocorrelation
Summer0.64DW < D(L)Positive autocorrelation
Fall3.24DW > 4-D(L)Negative autocorrelation

The number of observations in each series is 9. Because of inspection of the data on beforehand one degree of freedom has been subtracted which corresponds to 8 instead of 9 “effective” observations. The model has one parameter in addition to the intercept. Then the 5% significance levels for the DW test are D(L) = 0.763 and D(U) = 1.332. The DW test is considered to have no significant autocorrelations if D(U) < DW < 4-D(U), indifferent if D(L) < DW < D(U) and 4-D(L) > DW > 4-D(U), positive autocorrelations if DW < D(L), and negative autocorrelations if DW > 4-D(L).