Answers are grounded in this paper and cite their sources.
The paper compares the power functions of the conventional tests of significance based on asymptotic theory with those of some alternative tests suggested by Dhrymes and Richardson and Rohr and also a test that was suggested earlier by Anderson and Rubin in a simple two equation model by means of Monte Carlo experiments.
No takes yet. Share an insight, caveat, or question.
G. S. Maddala (1974) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: