Can We Build Trust?
Increased globalization, and competition are creating more global teams in engineering design, new product development and other tasks that require multiple disciplines. These global, cross-functional teams require trust between team members to accomplish their goals, but they also make trust difficult to achieve.
Members of these teams usually belong to different professions, ethnic groups, nationalities, and organizations, which makes trust difficult. Team members have different cultures, goals and expectations and this makes trust difficult, too. Distributed team members must use some form of technology to communicate. This can make communication more difficult and less personal. This could also make it more difficult to develop shared understandings, upon which trust can develop. Misunderstandings that can damage trust are more likely to happen and in this situation, damaged trust is more difficult to repair.
Until recently, there was little agreement about the nature of trust and how it develops. Some theories propose that interpersonal trust is based upon shared social categories, roles, third party information, social rules, history of the relationship and the trustor's disposition. Although most of these theories have been tested individually, no model combining these theories has been tested yet.
These theories lead to questions such as: When these trust development factors are combined, which ones are the most influential in predicting trust in global, cross-functional teams? Are these factors consistent under all conditions? What risk factors that could lead to a failure of trust?
These questions are important because as teams become more diverse, more distributed and tackle more complex tasks, the risk of failure due to lack of trust increases. When designing such teams, it would be useful to know if there are limits on the extent of team diversity that one team can handle, for example. Other factors that may also require team design constraints or tradeoffs are geographical distance, temporal distance and task complexity.
To tackle these questions, we integrated the theories of trust development
into a model that identifies the major variables that predict interpersonal
trust and describes the relationship between them. We then observed teams
at work in an educational environment and developed ways to measure the
variables in the model. For example, we measure trust by asking how much
a team member feels the need to check up on another
After three years, we have learned some interesting things about trust. New relationships usually start out with some level of trust. At the beginning of a relationship, the individual's perception of the risks and rewards of trusting is important in determining trust. In a mature relationship, trust is predicted by perceived trustworthiness and follow through, i.e. doing what was expected. Over time, these factors have a tendency to remain stable, but events that provide information about follow through, like deadlines, can influence trust.
We also found some interesting differences between distributed and collocate dyads. In collocated dyads, a team member's evaluation of another's follow through determines the perceived trustworthiness of the other party and that predicts trust. In distributed dyads, it is the other way around. The team member uses perceived trustworthiness of the other team member to evaluate the extent to which the team member is perceived to have followed through and that predicts trust. This implies that in a distributed team, team members use their opinion of the person to assess his or her performance. This could happen because in a distributed team it is difficult to get information about a team member's performance. In a distributed team that is also cross-functional, this difficulty could be accentuated by the inability to understand the other partner's discipline.
Now that we have validated this model of trust, we can apply it to other situations to see where the situations are the same or how they are different. We are currently testing the model in a real work teams in the construction industry. This will help us to extend the model, learn more about trust in different contexts and build trust in distributed, cross-functional teams.
|Modified 15 January 2003 * Contact Us|