Matter Over Mind? E-mail Data and the Measurement of Social Networks

Abstract

Organizational network scholars have not yet fully exploited the information revolution for data on intraorganizational social networks. To encourage research using electronic data, we analyze the correspondence between e-mail and survey measures of the same social network. Substantively, we find that clustering is explained by actor attributes (hierarchy, tenure and group) in the survey measure, but appears to be endogenous in the e-mail measure – that is, relative to electronic traces of observable interactions, survey respondents tend to over-state ties to high-status alters and under-state ties to physically and organizationally distant alters. We conclude that survey data provide information about actors’ perceptions of a network and should be used when those perceptions are of substantive interest. In contrast, observational data such as e-mails measure the objective communication structure and are a better data source for research questions that depend on measurement of the actual flow of communications. The authors would like to thank Galina Daraganova, Dean Lusher, Mikolaj Jan Piskorski, Pip Pattison, Gary Robbins, Toby Stuart, Mike Tushman, Tom Valente, Peng Wang, seminar participants at the Harvard Business School, and conference attendees at INSNA’s SunBelt conference for their valuable comments and suggestions. Any remaining errors are our own. Correspondence concerning this article should be addressed to Eric Quintane, Institute of Management, University of Lugano – Lugano, Switzerland, [e-mail]; phone [phone].