[R] Bartlett test

2007-01-19 Thread Matthieu Mourroux
Bonjour, Je voudrais tester l'homoscédasdicité entre des groupes. J'aurais alors aimé savoir quelle était l'hypothèse nulle du test de bartlett. Merci pour votre aide. Matthieu. This e-mail has been scanned for all viruses

Re: [R] Bartlett test

2007-01-19 Thread Benilton Carvalho
The null hypothesis is that the variances do not differ across groups. The Bartlett test is sensitive to non-normality and that might lead you to consider something more robust (eg, Levene's test). b On Jan 19, 2007, at 9:04 AM, Matthieu Mourroux wrote: Bonjour, Je voudrais tester