0
$\begingroup$

I am modelling house prices against sales amount using a simple linear regression model.

My SPEC (Option in SAS) says IID (p-value > 0.05) but my DW (Option in SAS) says a strong 1st order autocorrelation (DW<2). How can this happen? I have included the result screenshotenter image description here

Thank you!

$\endgroup$
0
$\begingroup$

Durbin Watson test it's not applicable in small sample sizes. If your data set has more than p*10 (number of independent variables * 10) observations then you can consider it large. If your data set has less than p*5 observations than you can consider it small. In a case of small sample, DW test cannot be applied and in such a situation bootstrap procedure may be a solution to this problem.

Also you can check the table of Critical Values of the Durbin-Watson Statistic to see if they are correlated or not, because DW ~2 it's a rule of thumb, also DW ~1.6 value are used often.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.