Re: [Gretl-users] serial correlation test

On Thu, 2 Aug 2012, Pindar wrote: > in general the test on serial correlation using the Q-statistic has the null > hypothesis of "No significant autocorrelations" (up to the lags used for the > test, which here are 4) > Hence, each p-value indicates that there is no serial correlation since you > cannot reject the null. All fine. > The critical value is 0,710723 with alpha=0,05 - use GRETL > to verify that ;-) That's actually the value for alpha = 0.95. The 5 percent critical value for chi-square(4) is 9.48773, via gretl's statistical tables.

Allin Cottrell _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users