1 term
A statistical procedure for deciding whether a suspect result differs enough from the others to be rejected.