Science 20 July 2012:
Vol. 337 no. 6092 pp. 337-341

Identifying Influential and Susceptible Members of Social Networks

Sinan Aral, Dylan Walker | 1 Comments

A randomized experiment based on product adoption among Facebook friends identified trend setters and followers.

Add a new comment

These postings do not necessarily represent the views/opinions of Science.

Aral and Walker cleverly show that, in the marketplace,influential individuals tend not to be susceptible to influence from others, while susceptible individuals tend not to be influential. Additionally, they show that influence increases with age, while susceptibility decreases. These findings are valuable not just because they show how goods are consumed, but because they could help us understand how social interactions shape individual choices from politics, to lifestyle, to fashions. Our theoretical work causally explains these results and suggests that they hold true across social domains. Our models show that whenever individuals share ideas (beliefs, attitudes, etc.), the ideas that increase influence and decrease susceptibility in the individual spread more readily. This is because influential individuals effectively transmit the very ideas that make them influential, and because non-susceptible individuals cling to their ideas, including those that lower susceptibility, and transmit them repeatedly. For example, few young people consider wealth a top priority. But this idea spreads successfully because wealth makes one more influential (likely to convince others that wealth is important) and less susceptible to alternatives (unlikely to willingly stop being wealthy). Over time and with exposure to many ideas, individuals accumulate those that lower susceptibility and increase influence (the others won't be retained or transmitted, by their very nature) and so the individuals become ever more influential and less susceptible. In our example, more individuals value wealth with increasing age.

These models make predictions not tested by Aral and Walker, such as that highly educated individuals may be both influential and relatively susceptible. The models, however, do not consider individual characteristics such as gender and relationship status, which Aral and Walker found important. Thus their results highlight the possibility for fruitful collaboration between empirical and theoretical research of social dynamics and the forces that shape our lives.

Stefano Ghirlanda1,2, Alberto Acerbi2, Magnus Enquist2 1 Department of Psychology, Brooklyn College 2 Centre for the Study of Cultural Evolution, Stockholm University

A Acerbi, M Enquist, S Ghirlanda, Cultural evolution and individual development of openness and conservatism PNAS 106:18931-18935, 2009 A Acerbi, S Ghirlanda, M Enquist, The Logic of Fashion Cycles PLoS ONE 7:e32541, 2012

Submitted on Thu, 12/20/2012 - 09:48