Daniel Berend

Senior Academic

Ranking of weighted majority rules

Daniel Berend, Yuri Chernyavsky, Luba Sapir

A decision-making body may utilize a wide variety of different strategies when required to make a collective decision. In principle, we would like to use the most effective decision rule, that is, the rule yielding the highest probability of making the correct decision. However, in reality we often have to choose a decision rule out of some restricted family of rules. Therefore, it is important to be able to rank various families of rules. In this paper we consider three classes of decision rules: (i) balanced expert rules, (ii) the so-called single expert rules, and (iii) restricted majority rules. For the first two classes, we show that, as we deviate from the best rule in the family, the effectiveness of the decision rule decreases. For the last class, we obtain a very different phenomenon: any inner ranking is possible.

Publication language English
Pages 994-1006
Journal Journal of Applied Probability
Volume 45
Issue number 4
Publication status Published - 01.12.2008

Keywords

Balanced expert rule
Decision rule
Dichotomous choice model
Restricted majority rule
Weighted majority rule

ASJC Scopus subject areas

Statistics and Probability
General Mathematics
Statistics, Probability and Uncertainty
Access to Document
10.1239/jap/1231340229
Other files and links
Link to publication in Scopus