Additionally, the observer ratings were based on an overall team function impression, not individual evaluations.
The team as a whole could have improved considerably in aspects of shared mental model, even though individual members themselves did not show substantial improvements.