Abstract
Global teams frequently consist of language-based subgroups who put togethercomplementary information to achieve common goals. Previous research outlines atwo-step work communication flow in these teams. There are team meetings usinga required common language (i.e., English); in preparation for those meetings,people have subgroup conversations in their native languages. Workcommunication at team meetings is often less effective than in subgroupconversations. In the current study, we investigate the idea of leveragingmachine translation (MT) to facilitate global team meetings. We hypothesizethat exchanging subgroup conversation logs before a team meeting offerscontextual information that benefits teamwork at the meeting. MT can translatethese logs, which enables comprehension at a low cost. To test our hypothesis,we conducted a between-subjects experiment where twenty quartets ofparticipants performed a personnel selection task. Each quartet included twoEnglish native speakers (NS) and two non-native speakers (NNS) whose nativelanguage was Mandarin. All participants began the task with subgroupconversations in their native languages, then proceeded to team meetings inEnglish. We manipulated the exchange of subgroup conversation logs prior toteam meetings: with MT-mediated exchanges versus without. Analysis ofparticipants' subjective experience, task performance, and depth of discussionsas reflected through their conversational moves jointly indicates that teammeeting quality improved when there were MT-mediated exchanges of subgroupconversation logs as opposed to no exchanges. We conclude with reflections onwhen and how MT could be applied to enhance global teamwork across a languagebarrier.