The purpose of this dissertation is to examine the communication behaviors of online leaders, or those who influence other members of online communities in terms of triggering replies, sparking conversations and diffusing language. It also examines the influence of group attributes on leadership such as size and connectedness. It relies on roughly 500,000 messages from 33,450 participants across sixteen discussion groups from GOOGLE GROUPS that took place over a two-year period. It utilizes automated text analysis, social network analysis and hierarchical linear modeling to uncover the language and social behaviors of online leaders. The findings suggest that online leaders influence others through high communication activity, credibility, reciprocal social network behaviors, and the use of affective, assertive and linguistic diversity in their online messages. Brokering, in which users connect to those who are not connected to each other, is not a significant predictor, suggesting that transparency and accessibility in online environments reduce the advantages of serving as a broker. In addition, group attributes such as size and network density encourage the emergence of leaders. However, participation equality and group turnover do not affect these behaviors, which emphasize the unique context of online communities, which often show power-law distributions of participation and high attrition rates. Taken together, the findings extend existing theories of social influence found in communication studies and social psychology, and increase our theoretical understanding of online leadership. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]