Email may not reflect the social network. Preliminary Insights

Researchers have demonstrated that e-mail archives can be used as a good approximation of social ties (Tyler et. al.). This paper demonstrates that ties obtained by mining e-mails archives do not necessarily provide a complete and realistic approximation of interactions by other communication channels. The results of our project indicate that factors such as co-location and the nature of working relationships influence the preference for rich media like chat and face-to-face. We have collected the e-mailboxes of a sample of 25 students and researchers, representing 50% of the entire population of a university research department. Through an online questionnaire sent out once per week for 7 weeks, we collected quantitative data on the average frequency of communication using different media: face-to-face meetings, chat, and telephone. During the three months period under observation, 64% of our sample members worked in the main Campus building, 20% in a secondary building and 16% off-site. The comparison of the complete networks suggests that face-to-face and chat are always positively correlated. In particular, people who are co-located, who are peers and who are working on the same project are more likely to use face-to-face and chat. As suggested by previous social network studies (Haythornthwaite, Wellman), the use of rich media strongly predominated, while telephone was scarcely used, and chat was enthusiastically adopted as a supplementary way to face-to-face interaction.