This should not be surprising when we consider that the representation relationship between the citizenry as a whole and the random sample is only a many-to-many relationship. Given that the sample as a whole represents the entire population, it seems plausible to assume that the former can provide interesting information about the latter to non-participating citizens. However, this is true only for information that has been collectively selected by all members of the random sample and is therefore representative of the sample taken as a whole.

Conversely, since the nonparticipating citizens have not chosen any particular member of the sample, there is C Level Contact List individual representation relationship between them and a particular  or specific subset of that group (there is no one-to-one representation relationship). one, nor one to many). Thus they have no reason to expect greater congruence between their interests and values ​​and those of any particular subset of the random sample, and thus no good reason to trust a recommendation that has been selected by only a subset of the random sample.

Sample regardless of how large or small each subset is. Thus, the problem with asserting that nonparticipating citizens should trust the majority's recommendation is not that it rests on the assumption that the random sample is a reflection "of the people." On the contrary. Precisely, if the reflection assumption is true, it is not possible for all citizens to be likethe majority and not the minority of the sample. Therefore, that statement cannot justify a general recommendation to trust the majority of the selection. Even if we assume that the selection reflects the town as a whole, this says nothing about which subset of the sample is most congruent with the interests and values ​​of each nonparticipating citizen.