Student groups completed a rank-ordering task in a "low-structure" computerized meeting room where all group members had equal access to a shared computer with a large monitor. Strategies used by the groups to distribute control over the public monitor, determinants of which members took control, and the consequences of control strategies were examined. Groups adopted either a dedicated-scribe strategy, in which one member had control throughout the session, or a non-dedicated-scribe strategy, in which control of the public monitor passed among members. Groups with at least one member who had low proficiency with the technology were very likely to adopt a dedicated-scribe strategy. Social influence within the group, proficiency with the computer system, and gender predicted which group members would take control of the public monitor. The results suggest that a group's social structure may be altered by the use of low-structure computer support, depending on the distribution of technical proficiency in the group. Dedicated-scribe groups had marginally better task performance but reported less increase in satisfaction (over previous work together) than non-dedicated-scribe groups. The implications of this research for the design and use of group computer support are discussed.