Answers are grounded in this paper and cite their sources.
Summary 1. 1. Professor Edgeworth has recently suggested the use of medians, as an alternative to least squares (L.S.), for reducing observations involving two variables, on account of their superior facility. Opportunity is therefore taken to press once more the claims of the method of zero sum (Z.S.), which is even simpler.2. 2. The method is described again with special reference to two variables.3. 3. Some more examples are given, including those of Prof Edgeworth.4. 4. It is urged again that there is no reason to believe that Z.S. is necessarily inferior in accuracy to L.S., or5. 5. If it is inferior, that the difference is significant.6. 6. Some methods are discussed whereby it might be possible to decide experimentally whether Z.S. is inferior to L.S. Such applications of them as have been made do not indicate any difference between the two methods appreciable to experiment.
No takes yet. Share an insight, caveat, or question.
Norman Campbell (1924) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: