Large-scale mobilization of individuals across social networks is becoming increasingly prevalent in society. However, little is known about what affects the speed of social mobilization. Here we use a framed field experiment to identify and measure properties of individuals and their relationships that predict mobilization speed. We ran a global social mobilization contest and recorded personal traits of the participants and those they recruited. We studied the effects of ascribed traits (gender, age) and acquired traits (geography, and information source) on the speed of mobilization. We found that homophily, a preference for interacting with other individuals with similar traits, had a mixed role in social mobilization. Homophily was present for acquired traits, in which mobilization speed was faster when the recuiter and recruit had the same trait compared to different traits. In contrast, we did not find support for homophily for the ascribed traits. Instead, those traits had other, non-homophily effects: Females mobilized other females faster than males mobilized other males. Younger recruiters mobilized others faster, and older recruits mobilized slower. Recruits also mobilized faster when they first heard about the contest directly from the contest organization, and decreased in speed when hearing from less personal source types (e.g. family vs. media). These findings show that social mobilization includes dynamics that are unlike other, more passive forms of social activity propagation. These findings suggest relevant factors for engineering social mobilization tasks for increased speed.

Na, not available.

MI, myocardial infarction; CVD, cardiovascular disease; ACS, acute coronary syndrome; IHD, ischemic heart disease; NA, not available; MONICA, the study focus on multinational monitoring of trends and determinants in cardiovascular disease; Inter99, the study focus on effect on IHD incidence of individually tailored non-pharmacological intervention on lifestyle using a newly developed computer-based health educational tool; CGPS, Copenhagen General Population Study; CCHS, Copenhagen City Heart Study; CIHDS, Copenhagen Ischemic Heart Disease Study.aThe power calculation was performed using Quanto software http://hydra.usc.edu/gxe/.

Demographic and clinical characteristics of the study cohorts.

PHTS phenotype and demographic data of research participants.