Good morning

We are not quite sure about the interpretation of the Durbin Watson p-value.

According to the user`s guide “an approximate P-value for the null of no serial correlation (rho= 0) against the alternative of  rho> 0 may be available via the accessor $dwpval”.

However when using this Gretl option it apparently provides the left-tail probability, for example

Durbin-Watson = 1,38  p = 0,003

Durbin-Watson = 2,33919  p = 0,785174

Assuming this, the Durbin Watson two-tailed test (rho= 0 against rho different from zero) should consider the p value:

2p in case DW is lower than 2 

2(1-p) for DW higher than 2

We would be very grateful if you could confirm this 

Thanks and best regards