Developing methods for analysing team mental model data

In this chapter, new statistical methods for analysing and representing team mental models (TMMs) are developed and explained. Current methodologies used to assess TMMs either focus on the 'sharedness' of mental models amongst dyads rather than the team as a whole, or consider within-team agreement in terms of a comparison to often inappropriate referent sets. Instead, they should carefully and systematically investigate actual patterns of observed data provided by team members. The newly developed method presented here measures agreement amongst team members by exploring the distribution of team members' responses to individual paired comparison items with reference to a predetermined external cut-off threshold based on an interpretation of the response scale provided. Detailed analyses are presented of pair-wise comparison data from a single team drawn from an existing data set obtained from a previously conducted 3-year field-based study. This new approach can determine, for each paired comparison rating item, whether a team is in agreement or disagreement, or whether the team has polarised views. This method allows for a measure of TMM congruence, which is an overall measurement of team agreement on a common mental model. After identifying the common TMM (i.e., overlap of individual team members' mental models), this technique can provide data to be used to graphically represent the congruent TMM