| As organizations become more team oriented, research on teams continues to increase in importance. This study furthers research on teams by providing a measure that captures members' experiences of conversational learning spaces in teams. In this study, I report the process of developing, refining and validating the measure through confirmatory factor analysis and structural equation modeling. The 5-dimensional model (with two poles for each dimension) of Conversational Learning Space with (Baker, Jensen, & Kolb, 2002) forms the main theoretical foundation from which the items were developed and tested. This study empirically tests this model and its impact on group effectiveness, member satisfaction, and psychological safety. After revising the Conversational Space Inventory (CoSI) items following a pilot study (n = 118), the instrument was administered to a total of 341 individuals in 49 teams ranging from educational to work groups. Psychometric assessments were carried out at the individual level to determine the factor structure. Findings show that conversational learning spaces are a group level construct (ICCs ranged from .2 to .4; r wgs ranged from .76 to .96) with four dimensions: Diverging, Converging, Status and Recursiveness. The results show that teams with a strong single leader (Status) significantly (p < .000) and strongly (but negatively) affects all three dependent variables. This strong negative effect is an unexpected finding and is reason for further research using work groups and groups across different cultures/nationalities. This study also highlights the importance of Recursiveness as a unique dimension that captures an aspect not presently studied---the experience of whether individual interests and issues are addressed in a team. Each dimension was tested for effect using path analyses and as a construct, conversational learning spaces was tested using Structural Equation Modeling. Results show that both the dimensions and conversational learning space as a construct strongly and significantly affect the three dependent variables (p < .000). Using Conversational Learning Spaces as an approach would help team researchers measure team functioning in terms of Real and Ideal Spaces. Further research using work groups and groups in cross-cultural settings would also contribute to this stream of research. |
