Condon HICSS 2001
Summary
The authors investigate the effect of communication medium on turn-taking in conversation, in terms of strategies that participants in a conversation use to manage the same work. They compared face-to-face to synchronous computer mediated conversation to asynchronous computer mediated conversation. These different channels cause different communication and coordination strategies to emerge, measured as number of messages exchanged, the length of the messages exchanged, and the content of the messages. Participants in synchronous communications manage their interactions differently than those in asynchronous, due to the different demands the media place on message processing and decision tasks.
The authors performed experiments in which participants participated variously in face to face, synchronous electronic communication and asynchronous electronic communication in the context of two planning tasks. In all cases, all communication between conversation participants was recorded. Dyads participating in the synchronous computer-mediated interactions were broken into groups, each group using a version of the software that offered a different number of messages worth of conversation history: 4, 10 or 18 lines (p. 2). E-mail was used as the medium for the asynchronous task. The conversations of face-to-face participants were recorded and transcribed for analysis.
The authors analyzed the recorded conversations in terms of structure and function: discourse functions and turn and utterance sizes. Turn boundaries occur at speaker changes (p. 3). All conversations were divided into "utterance units" which are "single clauses with all complements and adjuncts, including sentential complements and subordinate clauses" (p. 3), and a single turn may contain many utterance units, especially in the asynchronous cases. Utterance units were coded according to the conversational function they served (greeting, suggests action, elaborates, agrees with suggestion, etc.), and the authors used the functional coding to identify common turn sequences.
Face-to-face interactions are characterized by long sequences of short turns following a stereotyped interaction sequence. Each turn consisting of one utterance.
Issues
Channel characteristics
face-to-face communication is synchronous, meaning that participants "interpret their partner's contributions while simultaneously planning their own next contributions". The simultaneous production of messages can cause signal degredation as participants talk over each other, and the participants must rely on their short term memories in order to remember what has been said (p. 1).
synchronous electronic communication (such as UNIX talk or chat) is similar to face to face in that it shares the simulaneity of responses, but avoid signal degredation and typically a history of what has been already said is available to participants (although that depends on the system).
asynchronous electronic communication (such as e-mail) involves a typically long time delay between responses.
