ABSTRACT: This study analyzes computer-monitored and self-reported electronic mail usage and network data collected over time from mentors and their summer interns at an R&D organization. Amount and network measures of E-mail usage were significantly associated with work and work familiarity networks, As time passed, interns communicated through E-mail more outside their formal mentor-intern relations. However, amount of E-mail use and most E-mail network measures (such as centrality) were not related to mentors' assessments of interns' performance several months later. An intriguing exception was how interns were located in the overall E-mail network. Surprisingly, overall, most forms of communication were negatively associated with performance ratings. These results imply that it is not necessarily how much one uses an E-mail system, but how the users are positioned in that system's structural context, that may affect R&D performance.
Key words and phrases: diffusion, electronic mail, implementation, network analysis, organizational computer-mediated communication, organizational structure, R&D networks, R&D performance