The majority map and the recursions of conformist transmission and of Galam's rule #
The probability that the majority of three independent samples, each of type 1 with
probability p, is of type 1 is p³ + 3p²(1 - p) = 3p² - 2p³: the majority map
(majorityMap). It is the expected one-round map of binary 3-Majority and of 2-Choices (the
binary median rule), a mean field in the sense of Dynamics.Kernel.MeanField (packages
plurality and median), and the deterministic recursion of two models of other fields (rows
three-majority and two-choices of the cross-field dictionary):
- Conformist transmission (Boyd and Richerson 1985, Eq. 7.1). A naive individual samples three
models and acquires a variant held by 3, 2, 1, 0 of them with probability
1,2/3 + D/3,1/3 - D/3,0(conformistTable). If the models are independent and each holds the variant with probabilityp, the variant is acquired with probabilityp + D p (1 - p)(2p - 1)(conformistMap,conformistMap_eq_expect). The table adopts the majority of the three models with probabilityDand copies one of them, drawn uniformly, otherwise (conformistTable_eq), soD = 0is unbiased transmission (conformistMap_zero) andD = 1gives the majority map (conformistMap_one). - Galam's majority rule with groups of three (Galam 2002, Eq. 4 with
a_3 = 1): the proportion of an opinion followsP ↦ P³ + 3P²(1 - P)(galamMap), the majority map (galamMap_eq).
The majority map #
The majority of three independent samples: the map is the probability that at least two of
three independent coins of bias p show heads.
The two types are symmetric: the other type follows the same map.
Conformist transmission #
Conformist transmission (Boyd and Richerson 1985, Eq. 7.1): the frequency p of a
variant becomes p + D p (1 - p)(2p - 1) in the next generation, D measuring the strength of
conformism.
Instances For
Boyd and Richerson's table of conformist transmission: a naive individual whose three models
include j holders of the variant acquires it with probability 0, 1/3 - D/3, 2/3 + D/3, 1
for j = 0, 1, 2, 3.
Instances For
The table adopts the majority with probability D and otherwise copies one model. With j
holders among the three models, the acquisition probability is D times the indicator that they
are a majority plus 1 - D times the probability j/3 that a model drawn uniformly among the
three holds the variant.
Boyd and Richerson's Eq. 7.1 from the table. If the three models are independent and
each holds the variant with probability p, so that the number of holders among them has the law
Bin(3, p), the variant is acquired with probability p + D p (1 - p)(2p - 1).
Conformist transmission is a mixture of unbiased transmission and of the majority map.
With D = 1, conformist transmission is the majority map p ↦ 3p² - 2p³.
With D = 0, conformist transmission is unbiased: the frequency is unchanged.