Content area
Full Text
About the Authors:
Stefan Wuchty
Current address: National Center for Biotechnology Information (NCBI), National Institutes of Health (NIH), Bethesda, Maryland, United States of America
Affiliation: Northwestern Institute on Complex Systems and Network Science (NICO), Northwestern University, Evanston, Illinois, United States of America
Brian Uzzi
* E-mail: [email protected]
Affiliations Northwestern Institute on Complex Systems and Network Science (NICO), Northwestern University, Evanston, Illinois, United States of America, Kellogg School of Management and McCormick School of Engineering, Northwestern University, Evanston, Illinois, United States of America
Introduction
Shifts in human communication have raised new questions about how role relationships -friendship and professional ties - are expressed in novel, electronic communication. Yet, as digital communication increasingly expands, research emphasizes that the object of ultimate interest is not the full set of contacts a person communicates with, but the identification of the social network of contacts where proprietary resources flow [1]. Current work attempting to define the social network within the flow of communication is based on the use of nodal demographic characteristics [2] to suppose the presence of likely ties or on flow thresholds for converting continuous email transmissions to binary yes/no friendships [3], [4], [5]. Despite the development of sophisticated tools [6] little empirical evidence exists on the strength of the correspondence between self-reported social ties and actual communication dynamics [1], [6], [7], [8], [9], [10], [11], suggesting that such knowledge could help advance research across disciplines [1], [2], [3], [4], [5], [6], [7], [8], [10], [11], [12], [13].
Here, we used self-reported human relations and email data from a typical professional services organization to investigate how email communication patterns map onto self-reported social network data. The significance of our approach lies in the ability to directly compare rare self-report and email data of the same sample population. Specifically, relative to previous work on managers' networks [11], [12], [14], [15], [16], [17], [18], our contributions include (1) the unique opportunity to compare self-reported social network data and email derived networks. (2) Highly detailed self-reported data. Our data allows for respondents to list up to 9 contacts whereas other work permits only 1–3 contacts (i.e., General Social Survey), and our respondents specified contacts as professional, social, or mentor ties for finer grained distinctions than normally permitted. (3) Also extending recent...