Extension To Dependent Test Statistics
In cases where the tests are not independent, the null distribution of X2 is more complicated. A common strategy is to approximate the null distribution with a scaled χ2-distribution random variable. Different approaches may be used depending on whether or not the covariance between the different p-values is known.
Brown's method is designed to combine dependent p-values with known covariance while Kost's method is designed to combine dependent p-values with unknown covariance.
Read more about this topic: Fisher's Method
Famous quotes containing the words extension, dependent, test and/or statistics:
“We know then the existence and nature of the finite, because we also are finite and have extension. We know the existence of the infinite and are ignorant of its nature, because it has extension like us, but not limits like us. But we know neither the existence nor the nature of God, because he has neither extension nor limits.”
—Blaise Pascal (1623–1662)
“It is possible to make friends with our children—but probably not while they are children.... Friendship is a relationship of mutual dependence-interdependence. A family is a relationship in which some of the participants are dependent on others. It is the job of parents to provide for their children. It is not appropriate for adults to enter into parenthood recognizing they have made a decision to accept dependents and then try to pretend that their children are not dependent on them.”
—Donald C. Medeiros (20th century)
“We do not quite say that the new is more valuable because it fits in; but its fitting in is a test of its value—a test, it is true, which can only be slowly and cautiously applied, for we are none of us infallible judges of conformity.”
—T.S. (Thomas Stearns)
“He uses statistics as a drunken man uses lamp-posts—for support rather than illumination.”
—Andrew Lang (1844–1912)