An Expanded Representation of Belief Structures in Groups

In this paper we propose an expanded representation of group belief structures. Three key aspects of belief structures are identified – validity, irrelevance, and uncertainty. Validity arises from the amount of valid beliefs that comprise a group’s belief structure, and the amount of invalid (not corresponding to reality) beliefs in the belief structure. Irrelevance represents the represents the reliance of a group on beliefs that, while valid, do not significantly influence the group’s decision making ability. Finally, uncertainty represents the extent to which a group has confidence in its beliefs, or whether it simultaneously adheres to contradictory beliefs. The three characteristics of a group’s belief structure are show to significantly impact group short and long run performance. Beliefs held by groups are important constituents of their mental models (Shope, 1983; Smith, Benson, & Curley, 1991). They are also constituent elements of group schemas (Rousseau, 1985). Consequently, group beliefs are likely to shape the group’s approach to solving critical problems and determine the quality of their decisions. Beliefs are closely tied to knowledge – Researchers have pointed out that knowledge is comprised of ‘true justified beliefs’ (Nonaka & Takeuchi, 1995; Polanyi, 1958) where beliefs are considered to be true justified if they correspond with reality and if they have been formed through valid observation or through reasoned logic (Smith et al., 1991). For the most part, organizational researchers have focused on the true justified beliefs while ignoring other beliefs that may be held by the group. In reality however, groups also work with unjustified beliefs (for instance a marketing team may believe that their customers are insensitive to price while in reality they are not) or irrelevant beliefs (for instance, a product development team may decide on design criteria, which while having valid content, are unlikely to impact product performance or demand). A group will likely take action on the basis of their beliefs – whether those beliefs are valid, invalid, or irrelevant. Consequently an understanding the group decision making process requires us to consider all three types of beliefs simultaneously. * Dept. of Management, Sam M. Walton College of Business, University of Arkansas, Fayetteville, AR 72701, Phone: 479-575-4007 Proceedings of OLKC 2007 – “Learning Fusion” 19 In this paper we propose an expanded representation of a group’s belief structures. We discuss how groups may differ with respect to the nature of their belief structures, and the implications of different belief structures for performance. 1 INDIVIDUAL BELIEFS In this section, we briefly describe individual beliefs and then move on to explore belief structures at group levels. Individual Beliefs When attempting to represent an individual’s mental and cognitive states, psychologists have described two key states: a) Qualia which are qualitative states corresponding to feelings and emotions, and b) Propositional attitudes that represent ‘meaning or content that can be true or false (Smith, Benson, and Curley, 1991: p 292).’ While we can mentally form numerous propositions in our mind, when we gain confidence that a given proposition accurately represents reality, then that proposition represents our belief (Shope, 1983). Note that our perception that the proposition represents reality may or may not be true. Beliefs are inseparably intertwined with human knowledge. Polanyi (1966) describes knowledge as being composed of individuals’ justified true beliefs. Shope (1983) points out that the ‘true’ aspect of a belief implies that it corresponds with reality, while the ‘justified’ part implies that the belief has been derived from valid evidence, and/or by strong reasoned argument. (Thus a true belief that is held by fluke is not considered knowledge). In addition to justified true belief, individuals also act on the basis of other beliefs that are incorrect. Thus a marketing manager may strongly hold a belief that price reductions are strongly related to increased sales, but this may just be a wrong belief. Such a belief is a deviation from reality, but is important because it shapes the actions taken by the individual. Such beliefs are a neglected but important aspect on individual belief structures. In addition to knowledge and distortion, individuals may also possess irrelevant beliefs. These are beliefs that are true as far as their content goes, but are not essential for perform the task. For instance, old-timers in paper manufacturing mills believe that the ‘formation’ of the paper is an important determinant of the quality of paper. That is true, but with the emergence of advanced finishing and printing machines, formation was not an important determining characteristic for customer choice. Consequently, an individual selecting a new manufacturing technology while possessing the irrelevant belief may spend more time and actually make a sub-optimal choice. While the preceding discussion provides a clean differentiation between propositions and beliefs, and between justified, distorted, and irrelevant beliefs, the actual reality involves additional nuances relating to the confidence with which individuals hold beliefs. Individuals can hold different confidence levels about how well a given proposition represents reality. Thus, it becomes important to focus on the strengths of individual beliefs – a strong belief is one where the individual is fairly confident of its correspondence with reality. In the case of a weaker belief, an individual may believe that his or her belief probably corresponds with reality but is a lot less sure. Proceedings of OLKC 2007 – “Learning Fusion” 20 Further complicating the above situation is the fact that individuals can simultaneously hold contradictory propositions (Goldman, 1986). For instance a marketing manager may simultaneous assign a positive and negative relationship between the price of a product and its consumer demand. When individuals hold competing propositions that are assigned similar probabilities they experience uncertainty. When the need arises to resolve such uncertainty, the individual may try to gain confidence about one of the two competing propositions through introspection, experimentation, or information seeking. The preceding discussion suggests that an individual’s belief structure may be categorized along the following three parameters: a) Whether the belief is justified, distorted, or irrelevant, b) the level of confidence with which beliefs are held, and c) the proportion of beliefs that compete with other contradictory beliefs. These three characteristics are also relevant and underlie important characteristics of a group’s belief structure as described below. 2 THE STRUCTURE AND CHARACTERISTICS OF GROUP BELIEFS Individual group members bring their belief structures into a group. Groups then go through a dynamic process where individual knowledge and beliefs are shared, debated, and typically yield a group mental model. For instance, Gibson (2001) suggested that decision-making groups typically undergo a process comprising four stages: a) Accumulation: Where they acquire needed information about the task, b) Interaction: Where group members retrieve information and beliefs held by each other, c) Examination: where group members examine each others beliefs, engage in interpretive discussions, and try to resolve inconsistencies and contradictions among each other’s beliefs, and d) Accomodation: where group members integrate diverse knowledge of individual group members and develop a common team mental model. In a group when individuals interact with each other, the group begins to form a shared mental model and the group’s belief system emerges (Gibson, 2001; Levesque, Wilson, & Wholey, 2001; Mohammed & Dumville, 2001). During this process it is necessary to distinguish between three sets of group beliefs. The first set, that can be termed a “potential belief set”, comprises a simple pooling of all the beliefs that are held by group members (Anand, Fugate, & Manz, 2000). However, research in group information sharing has shown that decision-making groups are not likely to share all personal knowledge held by members and further, they are unlikely to use all of the knowledge that they do share (Anand, Glick, & Manz, 2002; Anand, Manz, & Glick, 1998; Stasser & Stewart, 1991; Stasser & Titus, 1985). Thus, at least two sub-sets of the potential belief sets are identified : The shared belief set, which comprises those beliefs that members have chosen to share with other group members, and an ‘active belief set’ which represents those beliefs that form the basis of a group’s actions. The relationship between the three group sets is shown graphically in Figure 1 below. The potential belief set is an artifact that is constructed by pooling together all beliefs that are held by group members. Individual members choose to share a portion of their beliefs with the group – their choices determine the nature and amount of information available in the shared belief set. The group’s active belief set, on the other hand, is a group level phenomenon and comprises those beliefs from the shared set that the group has Proceedings of OLKC 2007 – “Learning Fusion” 21 incorporated in its active memory or in its mental model (Anand et al., 2000; Mohammed et al., 2001). Past research has shown that the relationship between the potential belief set and the shared belief sets is likely to be moderated by group characteristics such as group diversity, amount of shared knowledge and so on. (Larson, Sargis, & Bauman, 2004; Larson, FosterFishman, & Keys, 1994; Stasser, Stewart, & Wittenbaum, 1995; Tesluk & Mathieu, 1999). Similarly, the relationship between the shared and active belief sets is likely to be moderated by the group characteristics and the process used by the group during discussions (Stasser, Vaughan, & Stewart, 2000; Stewart & Stasser, 1995). The active belief set is not necessarily a smaller sub-set of the shared belief set. The sharing of beliefs often yields new insights and may lead to the generation of specific beliefs that were not contained in the pooled or shared belief sets (Weick, 1995). While the relationships between pooled, shared, and active belief sets are important in themselves, they are beyond the scope of this paper. Rather, we begin with the assumption that groups have an active set of beliefs that guide their actions. The active set of beliefs can be viewed as a component part of a team’s mental model which is defined as the team members’ shared, organized understanding and mental representation of knowledge about key elements of the team’s task environment . The beliefs held in a group’s active belief set are similar to those held by individuals in that they comprise propositions that the group believes correspond to reality (to a greater or lesser extent). A representation of group belief structures is shown in Figure 2 below. Figure 1: From Individual Beliefs to Group Beliefs