1
Introduction
Information Exchange in Decision-Making Groups and Teams
Overview
This chapter introduces an information exchange framework for the study of decision making by groups and teams. While decision making by interactive units has been given recent impetus by technology that supports virtual teams, fundamental issues in the efficacy of interactive groups and teams remain. The differences between inferences from lab-based experimental studies and studies of organizational teams indicate some of the basic issues. Following a review of the results of a range of studies, it is suggested that the social structure of groups and teams is a source of process losses in these decision-making units that has not been adequately recognized or defined in available studies. Extensive background results in the study of group processes in support of this claim are cited. An account of ill-structured decision making as information exchange is then introduced. In this account, it is recognized that information exchange is in part a social process that introduces social risk to group and team members. This risk occurs because the receipt of negative evaluations can be a meaningful source of status loss for members. The magnitude of the loss can be expected to depend on judgments of the status distance between the source and recipient of the evaluation. Subsequent chapters that offer analytical, numerical, and empirical support for the information exchange perspective that is introduced are briefly reviewed.
Introduction
The ascendency of decision making and problem solving by interactive groups and teams is increasingly evident in contemporary organizations (Beers, Boshuizen, Kirschner, & Gijselaers, 2006; Cross, Thomas, & Light, 2009; Park & DeShon, 2010; Rowland & Parry, 2009) as are reports of their effectiveness (e.g., Feri, Irlenbusch, & Sutter, 2010). This is most evident in cases where decisions or problems cannot be given algorithmic forms or even well-defined heuristic forms. Following others (e.g., Mintzberg, 1973; Walker & Cox, 2006), this case will be referred to as ill-structured. Putative reasons for the use of interactive groups and teams as units in ill-structured decision making include the diversity of expertise and ideation that can be exchanged between decision makers (Koh, 2008), increases in perceived equity that participation can impart (Deutsch, 2010), and the contribution that participation in decision making can have on implementation of the decision (Tegarden, Sarason, Childers, & Hatfield, 2005).
However, the inferences of those who study the effectiveness of interactive teams in organizations (e.g., Malhotra & Majchrzak, 2005) differ from the most common conclusions of those who have studied interactive groups in controlled experiments (e.g., Kerr & Murthy, 2004; Paulus, Putman, Dugosh, Dzindolet, & Coskun, 2011; Pinsonneault, Barki, Gallupe, & Hopper, 1999; Strobe & Diehl, 1994). Reports from both contextualized implementation and controlled studies contribute to our understanding of process and efficacy of interactive groups and teams as decision-making units. Reconciling these differences is consequently important to the continuing use of groups and teams as decision-making units. In the exposition to follow, inferences from both these sources is assessed to further a more comprehensive conceptual account and its testing and contribute to the design of effective decision-making groups and teams.
Information Exchange in Decision Making
The perspective of the exposition is one in which microprocessing in decision making of interactive groups and teams is conceptualized in terms of information exchange. The quality of a group or team decision will be considered in terms of the amount and type of information that is exchanged and its sequencing. While understanding microprocessing is fundamental in assessing interactive decision making, recognizing social processes in agent interaction within groups or teams increase the complexity of accomplishing this. Team members as individual agents clearly have internalized personal objectives. However, when agents become team members, they in some part assume the objectives of the team. At the least, this requires an integration of agent objectives with those of the aggregate.
In considering processing in information exchange in interactive groups and teams, structure in the unit is introduced as an organizer of the content and source of information that is exchanged. Group or team structure is defined here in terms of the distribution of member status in the unit.
Social Structure in Decision-Making Groups and Teams
As well recognized, a status organization is generally emergent in interactive groups and teams (Berger, Fisek, Norman, & Zelditch, 1977; Berger, Rosenholtz, & Zelditch, 1980; Berger & Webster, 2006). The basis to expect structure to have pervasive effects on the amount, type, and distribution of information inputs across team members is elaborated.
A range of background studies indicate the stability and performance consequences of social structure (e.g., Parise & Rollag, 2010; Sperber & Hirschfeld, 2004). Elucidating how the social structure of groups and teams can affect the organization of information exchange and, consequently, the quality of decisions will be a principal objective of this exposition. Appropriate attention will be given to technology enabled methods to manage in social groups and team structure and information exchange toward quality objectives in decision making.
In the rest of this chapter, I introduce major contentions of the conceptual framework and provide brief overviews of the chapters that make up the discourse on decision making by groups and teams. In the next chapter, I elaborate on the history and background of fundamental processes in the interactive groups and teams with task-directed objectives. This includes the basis for linking the amount and type of information that members exchange to both decision quality and social structure in decision-making units.
Pervasiveness of Group Processes across Decision-Making Contexts: R&D Teams and Juries
The importance of social structure has now been indicated in a diversity of contexts that include juries and R&D teams. Teams in both of these applications are assumed to be most effective when they are putatively objective and egalitarian. There is now considerable evidence of the basis that social structure introduces in both these applications of interactive groups and teams. For example, Cohen and Zhou (1991) examined the effects of team, organizational, and external (societal) status characteristics on interaction patterns of established work teams. Both status characteristics that are external to the team and status characteristics within the team were shown to influence team interaction.
When status within the team is controlled, only one of the external characteristics has a direct effect on interaction. However, status within the team was found to be significantly influenced by each of the putatively external characteristics. These results indicate that competence and performance are clearly not the singular bases for team status. As such, they suggest that status processes in teams closely follow those reported in zero history lab groups. That is, beliefs that follow from diffuse status characteristics commonly affect the social order in interactive groups and teams.
In the case of juries, Davis and collaborators (Davis, Hulbert, Au, Chen, & Zarnoth, 1997; Kirchler & Davis, 1986) have followed the historical study of Strodtbeck and Mann (1956) and extensively examined group processes in their interaction. (Also see, Rashotte & Smith-Lovin, 1997). Most recently, Salerno and Diamond (2010) have reviewed classic jury decision-making research on jury deliberation that challenges the view that deliberation itself does not have an important effect on verdicts.
These authors call attention to cognitive processing during deliberation that might explain the transition between predeliberation predictions and a juryâs ultimate verdict. Diamond, Rose, Murphy, and Meixner (2011) have subsequently reported direct studies of real jury deliberation and awards that show damage anchors in awards to defendantsâ applications to follow from cognitive processes. Although not directly cited, there is a basis to expect that the processing of information described in these studies is organized by structure as it is in other contexts.
Following informal reports and group studies that have been cited, status organizing process can be expected to be operative in cognitive processing when jury members interact in deliberation. The present exposition defines and gives forms to the microprocessing of information that result from status orders and consequences for decision making in interactive groups and teams that are applicable across a range of contextual objectives.
Ill-Structured Decision Making
The early history of task-directed groups (e.g., Bales, 1950) directed attention to problem solving without defining the process or offering a distinction from decision making. The case of ill-structured decision making as defined above has a commonality with general designations of problem solving. Both cases do not typically begin with a set of well-defined alternatives as in discrete choice models of decision making. Additionally, they have a common dependence on the exchange of ideas and their evaluation. An important difference between decision making and problem solving is that the designation of decision making does commonly imply a definition of discrete alternatives and convergence to one of the alternatives. This is typically what teams in organizations are charged with doing and will be the designated process for which objectives are defined here.
For firms, the case of ill-structured decisions includes a range of critical decisions that range from the selection among strategic alternatives and the retention of new products (e.g., Schmidt, Montoya-Weiss, & Massen, 2001). This case clearly requires a more comprehensive consideration of the objectives of decision-making units and member contributions to these objectives than do applications in brainstorming traditions. In these traditions, the objective is strictly defined as maximizing the number of ideas generated by the group or team. Decision-making objectives increase the importance of evaluations as an information type. Additionally, the sequential exchange of information can consequently be expected to be more complex in microprocessing than in studies of problem solving. The design of systems that support group decision making (DSSs: e.g., Gallupe & DeSanctis, 1987; Holsapple & Whinston, 2000) correspondingly require a commensurate representation of the complexity in microprocessing. In the exposition to follow, this requires dynamic forms for information typologies in the presence of social structure. The proposed account of microprocessing in ill-structured group decision making is used to provide guidance for the design and implementation of GDSSs.
Quality in Group and Team Decision Making
As reviewed, the preponderance of evidence from experimental studies of brainstorming indicates that nominal groups are more productive than interactive groups. While this dialogue and its evidence is not the principal interest in the discourse to follow, it remains relevant to understanding and designing decision-making teams. In agreement with these studies, it is recognized that when design does not effectively manage the challenge of integrating microprocessing of agents as both individuals and group or team members, there is commonly a net loss in the quality of ill-structured decisions from members interaction in comparison to the quality offered by members working independently. Although the formal evidence remains ambiguous at present, it is suggested that effective understanding and management of processing in interactive groups and teams can result in contributions to decision quality that significantly exceed those offered by the same numbers of individuals working independently.
The intention of the exposition to follow is to propose an organizing framework for decision-making groups and teams that integrates process that is generated by both the objectives of members as individual agents and the commitment to group objectives. This will be seen as necessarily requiring the representation of the position in the social structure of the decision-making unit that an agent has. The above is clearly a complex agenda that benefits from taking note of extensive histories in the study of interactive groups. In the next sections of this chapter, I first indicate the pervasiveness of the structural effects in interactive groups and teams that have been cited. I then discuss the basis of distinctions between decision-making units as groups and teams and the decision typology that follows. Finally, I review relevant histories in the study of interactive groups as background for the exposition.
Designations of Groups and Teams
The designations of groups and teams are often used coordinately. I next briefly take note of concordances and divergences in the designations of groups and teams that are relevant to the present exposition. One of the more available discussions of teams has cited differences from groups in that the former evidence of greater commitment and synergy among members, more egalitarian interaction and longer existence (Katzenbach & Smith, 1993). However, disjunctive designations of groups and teams often too narrowly define groups and too quickly dismiss limitations in teams that are likely to correspond to those commonly reported in interactive groups. It is notable that many of the now cited differences between groups and teams were identified in earlier academic discourse on differences between effective and ineffective work groups (e.g., Campion, Medsker, & Higgs, 1993).1
While delineation of group and team distinctions are not exact, the principal interest in this exposition is in what are now discussed as teams. Since the preponderance of background laboratory studies use the designation of interac...