Hi,

is there a similar rule for the tolerable serial correlation/ auto-correlation by a regression analysis like the maximum-minimum quotient criterion for tolerable variance inhomogenity?