# DW Statistic: Decision rule

Somehow quite confused about the decision rule. Could someone give a brief?

If DW stat < DW Lower ==> reject, there is positive serial corr

DW lower < DW stat < DW upper ==> inclusive

DW upper < DW stat ==> do not reject, no evidence of positive serial corr.

wondering what is the different with inconclusive with do not reject, both mean no evidence of positive serial correlation, isn’t it?

You’d better take a look at Schweser 2011, they have a more detailed chart where we have d_u, d_l, 4-d_u, 4-d_l, like if DW > 4-d_l we have negative correlation.

generally you’re safe to assume a DW stat around two equals no serial serial correlation, low DW - positive serial corr and DW significantly greater than two = negative serial corr. i’m terrible at quant but this general guideline has allowed me to answer most questions regarding Durbin.

Everyone,

Good luck to all of us tomorrow, don’t forget: time management is the key to success.

Everyone,

Good luck to all of us tomorrow, don’t forget: time management is the key to success.

Everyone,

Good luck to all of us tomorrow, don’t forget: time management is the key to success.

cvv