An application of the socioegocentric model to information-sharing discussions: In search of group-level communication influences

Hewes's socioegocentric model is applied to discussion data from an information-sharing task. The primary question is whether discussion data display psychological (individual-level) or interactional (group-level) associations among contributions. Two studies are described. The first used a card sort task to group the 42 items from the choose-the-best-candidate task. The second study used the groupings from the first task to code information-based contributions to discussion. Results indicated that participants reliably group the items into valence-based categories and that there were valence-based associations between items contributed to discussion. However, the associations were at the individual level of analysis-there were no group-level effects for information sharing. However, other features of the discussions did have group-level effects, and we argue for a better understanding of information sharing in terms of the comments that frame information-based contributions to interaction.