2020 Theses Doctoral

Patterns of Intellectual Collaboration and Effectiveness of Small Groups

Lai, Huei-yi

The present study examines the collaborative behaviors of 35 students of ages ranging from the mid-20s to early 40s (19 male and 16 female) enrolled in an intensive one-week graduate business course as they engaged in a competitive strategic decision-making simulation. Each group represented a firm. Their discourse in small groups of 3-4 was audio-recorded as they made a series of decisions over the course of the week.

The central research question is whether the analysis of audio recordings of small group interactions would reveal any cognitive features of the small group collaboration that would be predictive of a group's performance (measured by the final stock price of the group's firm at the end of the simulation). Discourse transcripts were analyzed using a relational and functional coding scheme that classified each utterance expressed during group interaction with respect to its function in relation to the utterance preceding it. Chosen for analysis was both an early and a late session since early tasks, such as establishing shared understandings, may entail different processes and patterns of interaction than later ones, such as reaching joint conclusions. Particular attention is focused on meta-level utterances, defined as those statements that reflect on the activity, rather than constituting a part of the activity itself and addressing the task subject matter. These meta-level statements are further divided into Meta-Self and Meta-Group categories, i.e., those statements that refer only to the self's thinking and those that refer to the group. A hypothesis was that only Meta-Group statements would benefit coordinated action and hence, group performance.

The analysis focused on a comparison of the discourse characteristics of the highest and lowest performing groups. Results showed that members of the highest-performing group made significantly more Meta-Group utterances. In contrast, members of the lowest-performing group made more Meta-Self utterances, in proportion to total utterances made by each group. This difference was consistent across time. Examining the full sample of ten small groups, the pattern was similar. Results showed associations with performance outcomes for both of the discourse categories of major interest – most prominently for the frequency of Meta-Group utterances and to a lesser (and inverse) extent for Meta-Self utterances. The possibility is considered that Meta-Group discourse is productive because it represents a form of perspective-taking with respect to the group rather than the individual. Discussion of what the present data suggest about group process is supported by excerpts from individual post-course interviews and selected small-group discussions.