Cognitive Model of Team Performance

The drawbacks of fi HRA are attributable to its basic assumption of the decomposition principle that a human task can be decomposed into elementary task units. It is equivalent to the assumption in the linear system that the whole is the sum of its parts. It will be shown in this subsection that this assumption does not apply to team performance. Since teamwork is used in most business settings, the reliability of team performance must be assessed in PRA, and some model of team performance is required to do so. The simplest approach is to combine multiple models of individual performance and this approach was actually taken in the early stage of development. A team, however, is a nonlinear system so that team performance is greater than the simple sum of individual performance.

The cognitive processes of team performance can be effectively described by the concept of mutual beliefs. Tuomela and Miller introduced a notion of “We-Intentions” to describe the cognitive mechanism in a cooperating team as follows [11]. When a team composed of two members, A and B, intends to do a cooperative task X, the following conditions hold.

(1) A/B intends to do A's/B's own part of X. (intention)

(2) A/B believes that B/A will do B's/A's part of X. (belief)

(3) A/B believes that B/A believes that A/B will do A's/B's own part of X. (belief on belief)

Beliefs like (2) and (3) in the above, which can be recursively defined, are called mutual beliefs. Such an explanation of the cooperation mechanism using one's own cognitive state and a corresponding structure of recursive beliefs can clarify the constitutive meaning of “sharing” intentions by cooperating team members.

Kanno applied the above notion of mutual beliefs not only to team intentions but also to cognitive team processes in general and proposed the Mutual Belief Model (MBM) to represent the team cooperation mechanism [12]. Figure 24.4 shows a recursive structure of cognition and corresponding beliefs of a two-member team. The recursive structure of mutual beliefs can be theoretically defined ad infinitum, but the three layers shown here will be sufficient to describe realistic cooperating situations.

One's own cognition on the state of the external world and oneself is described in the first MBM layer. The beliefs on the partner's cognition are described in the second MBM layer, which is a reflected image of the partner's first layer. The third MBM layer is for describing the beliefs on the partner's beliefs on one's own cognition. It is one's self image through the partner. Since the second and the third MBM layers are nonexistent in the cognitive model of an individual, a model that merely combines individuals will not contain both layers.

Cooperative team performance can be achieved using all of these MBM layers. Cognitive entities on each MBM layer are obtained and related by various types

Fig. 24.4 Mutual Belief Model and interactions

of interactions within the layer or between different layers. These interactions are classified into four types: verbal communication, mental simulation, complementing, and verification.

Verbal communication is a process to transfer some cognitive entity from one person to another by explicit utterance. Mental simulation is a process to derive new cognitive entities from some others within the same MBM layer by inference using knowledge and manipulating mental models. Mental simulation is a process for interpretation and prediction not only of the state of external world but also of the partner's behavior. In complementing, some cognitive entity will be copied from one MBM layer to another within the same person. One adopts this scheme, for instance, in an occasion where he/she supposes his/her partner believes X because he/she believes X. Such a supposition, however, sometimes results in false presumption. Finally, verifi is the comparison of cognitive entities between different MBM layers to check consistency among mutual beliefs.

The cognitive processes mentioned above are nonlinear effects in terms of a combination of individual cognitive processes, and MBM becomes much more complex for a team larger than a dyad. Team cooperation by humans is more than simple division of labor. Accidents often occur with highly automated systems with no hardware failures, because mutual beliefs and cooperating interactions are lacking in systems where a linear human-machine combination is assumed. Consideration of the nonlinear nature of team performance is necessary also for sophisticated human-machine cooperation.

Safety Culture and High Reliability Organization

Safety culture was a new concept in systems safety that was introduced after the Chernobyl accident. As already mentioned, many organizational and social factors were found behind the direct cause of the accident, the operators' violation of the operation rules. This fi led safety specialists to attend to safety culture. Safety culture resides at the basis of the three factors shown in Fig. 24.3 that form the context of human performance. In order to prevent organizational accidents, safety culture has to be implemented and maintained by organizations.

A key question is how we can implement safety culture in organizations and maintain it. Research on organization science, in particular on high reliability organizations, gives us valuable implications to answer this question. A High Reliability Organization (HRO) is an organization where accidents and incidents are suppressed below the standard level of the related industry sector. The idea first came from the pioneering work by a group at the University of California, Berkeley [13]. This group examined behavioral patterns of work groups under high-risk and stressful conditions such as aircraft carriers, air traffic control, and nuclear power plants. From these studies, the characteristics observable in common among various HROs have been revealed, which is represented in a word, mindfulness. Mindfulness consists of the following five elementary characteristics:

• Preoccupation with failure;

• Reluctance to simplify interpretations;

• Sensitivity to operations;

• Commitment to resilience;

• Deference to expertise.

Organizations that incorporate the above characteristics can handle unanticipated situations skillfully and can recover from emergency rapidly.

Safety culture and HROs first drew attention for solving problems in the era of socio-technical interactions: how to establish proper interactions between technologies, organizations, and society, and how to avoid organizational accidents. These concepts, however, are related also to the ability of socio-technical systems to cope with unanticipated situations as suggested in the fourth item of the above list, and they give us implications for the era of resilience. A High Reliability Organization is sometimes characterized as a learning organization, the ability to adapt to changes and disturbances by restructuring itself is an essential requirement of a resilient system [14].

 
< Prev   CONTENTS   Next >