Homophily and the Speed of Social Mobilization: The Effect of Acquired and Ascribed Traits
Figure 1
Mobilized teams grew to a variety of sizes at a variety of rates.
(A) An example team growing from generations of recruiters to recruits, with different recruiter-recruit mobilizations having different types of links. The team starter's icon is black, and the future members decrease in shade as their generation in the team increases. Blue links indicate the recruiter and recruit heard about the contest through the same type of source (ex. friends). Red links indicate the recruiter and recruit heard through different types of sources (ex. family vs. the media). Green links indicate one or both participants did not give information on this personal trait. This example team was the 4th largest in the contest. (B–C) Using a similar social mobilization incentive system to that used in the present study, previous research suggested the distributions of team sizes and of recruiters' number of recruits followed power laws, with α of 1.96 and 1.69, respectively [12]. We used the statistical methods of Clauset et al. [31], [32] to find weak to modest support for discrete power laws on these metrics, though the power laws' scaling parameters α are replicated. Distribution plots are complementary cumulative distributions (survival functions). (B) Team size. There were 148 teams, with 51 recruiting additional members beyond the founder. The power law fit was preferred over an exponential (LLR: 58.53, p<.01), but was no better of a fit than a lognormal (LLR:.01, p>.9) (C) Number of recruits for each recruiter. There were 1,089 participants, with 152 mobilizing at least one recruit. The power law fit was better than that of an exponential (LLR: 61.45, p<.02), but was not a stronger fit than the lognormal distribution (LLR:−.04, p>.9)