A fixed alternative keeps a non-null data-generating distribution unchanged as sample size grows. A consistent test rejects with probability tending to one under every fixed alternative in its stated class.
New to topics? Read the docs here!
New to topics? Read the docs here!