Talk:Portmanteau test

From Wikipedia, the free encyclopedia

It would be nice for this article to at least contain the null hypothesis of the test. That's what I came here for. :-) StatsJunkie 03:03, 4 December 2006 (UTC)

The null hypothesis is "the data are random", which is confirmed if the test statistic is less than the critical value of the appropriate chi-squared distribution (i.e. if you want to test for significance at the 95% level and you're calculating the statistic up to lag 10, then you want to test it against the value p for which P(chisq(10) < p) = 0.95). That said, I was practically rewriting this article as "Ljung-Box Q" based on a listing at Wikipedia:Requested articles, so instead I'm going to redirect here and try to work out whether the statistic needs its own article or if I should just flesh this one out. Confusing Manifestation 06:12, 18 September 2007 (UTC)