Garold Stasser - US grants

When solving a problem or making a decision, small groups can potentially benefit from having members who can contribute a diversity of knowledge and insights. Previous research suggests, however, that decision.making groups (i.e., groups asked to make a judgment) often fail to consider information that is held by single members. Face.to.face, unstructured discussion focuses on information that all members already know, and, as a result, groups frequently overlook an option that is favored by their collective knowledge. The proposed research will examine the extent to which this phenomenon generalizes to groups that are solving a problem with a demonstratably correct answer. Additionally, the research will examine the role of strategically composed minority factions in facilitating more effective information exchange during group discussion. By tracking the flow and impact of information during discussion, the research will also provide data needed to refine the DISCUSS model of group decision making. This model simulates the sampling and use of information during face.to.face discussion en route to a group decision. It provides a convenient way of representing the complexities of group interaction and of examining the implications of changes in group composition (e.g., size and factional representation) and task demands (e.g., number of decision options or possible solutions and information load).

9410584 Stasser ABSTRACT Members of decision-making groups often have access to different information, and one aim of discussion is to pool members' unique knowledge. Pooling information is particularly important if the combination of members' unique information will reveal a superior decision alternative. However, previous work shows that discussions tend to focus on commonly-held knowledge, and groups risk overlooking an option supported by their collective knowledge. This program of empirical work examines how perceptions of knowledge distribution and task demands affect information processing and communication styles. Of particular interest is the process of tacit coordination whereby members of a group adjust their selection, rehearsal and communication of information to obtain varying degrees of informational diversity depending on perceived task demands. This coordination can be based on direct knowledge of others' access to information or on acquired knowledge of what others have done on similar tasks in the past. It is hypothesized that some types of tasks (collective recall, idea generation, intellective problems) will encourage diversity of informational input whereas other types (judgmental tasks) will promote uniformity of input. It is also hypothesized that concerns about recall accuracy may promote redundancy of informational sampling in order to provide social validation. It has been found that, in the process of decision making, groups tend to focus on the information that is commonly held by all members, ignoring information that is not shared by all but that could lead to a better decision. By advancing our understanding of the conditions under which unique information is likely to be shared, this research will provide knowledge relevant to improving the quality of decision making in groups. As such, it will have implications for various initiatives in industry and government that have emphasi zed the importance of basing decisions on the best information available from groups or teams. At the heart of these initiatives is the goal of empowering people who are uniquely aware of crucial information that could be used to improve processes and procedures, getting them to speak up and make a difference. The proposed research will help to provide a sound basis for such attempts to enhance worker productivity.

Much of the work that is accomplished in organizations, schools, families, and government settings involves people working together in groups. This project examines how it is that people are able to coordinate their efforts in cooperative, interdependent workgroups operating in multiple task environments. The research is informed by a model of tacit coordination, which suggests that individual group members allocate their efforts to tasks based on expectations about teammates' behavior (social knowledge), perceptions of task demands (task knowledge) and assessments of one's own capabilities (self knowledge). Theoretically relevant features of the task environment include group size, the number of tasks to be accomplished and the staffing demands of the tasks. The Tacit Coordination Model incorporates constructs from traditional decision theory and yields normative predictions for task choices based on the assumption that members choose tasks that maximize the marginal expected utility of their actions. Nonnormative heuristics for task choices are also considered. For example, members may avoid tasks for which their contributions are likely to be dispensable (overstaffed tasks) or futile (understaffed tasks). Six studies use a collective problem-solving paradigm and examine the development of coordination over repeated trials. These studies explore the use of normative principles and nonnormative heuristics by members to guide their task choices and the relationships among these rules and the effectiveness of team coordination. Also of interest is the impact of group heterogeneity (diversity of abilities and values) on coordinated action for tasks that have different staffing requirements. The ideas embodied in the model suggest, for example, that heterogeneity facilitates coordinated action when tasks can be staffed by one member but impedes coordination when tasks must be staffed by all members. The empirical work will inform the development of a computational model of tacit coordination, which can facilitate studying the collective implication of individual choices and the emergent properties of using different coordination rules over time.

The use of computer-supported interaction has become a primary feature of communication among group members, due in part to its structural (freedom of time and geographical constraints) and psychological (anonymity) features. As a consequence, many group researchers have investigated the role of group process variables in computer-supported interactions. Because groups communicate via the use of computers in many personal, educational, and professional settings, it is important to continue and encourage the study of group processes in such environments. One theoretical issue that is implicit in many of the studies of group process variables in computer-supported interaction is whether technology is deterministic or determined. Spears and Lea (1994) recognized this issue when discussing whether technology was a panacea, allowing for greater freedom of expression and equality of status, or a panopticon, limiting expression and increasing control due to its ability for surveillance. The purpose of this conference is to bring together a diverse group of researchers with established programs of research on group processes in computer-supported interaction. This conference should serve to make the theme of technological and social determinism more explicit, thus leading to greater unity in the dominant research programs. The proposed conference offers a unique and timely theme, and is cross-disciplinary and cross-cultural.

Small groups make important decisions in government, organizations, and education. Often the goal is to pool diverse information from group members and use it effectively in order to reach more informed decisions. The proposed work will examine the role of task framing in promoting more effective individual and collective decision-making. Many decision problems can be resolved by using heuristics or by applying systematic, analytical procedures. For example, one can evaluate prospective employees by judging their similarity to a prototypical "good" employee (heuristic frame). Conversely, one could use test scores, level of education, and GPA and other measures that are known to predict job performance (analytical frame). The proposed work examines the interplay of task framing at the individual and group level and its effects on decision-making.

This work will also explore different ways of distributing access to information that may facilitate problem-solving styles in collective decision-making. Preliminary work has shown that people who anticipated working in a group and thought that they had more information than others performed better on reasoning tasks. This work further suggests that the improved performance was partly due to "more informed" individuals processing information more deeply and framing the task in a more analytical way.

In a series of proposed laboratory studies, individuals and groups will complete social prediction tasks based on descriptive information about targets and base rate information that can be used to derive a prediction. The distributions of descriptive and base rate information among group members before discussions will be systematically varied. The descriptive information will be stereotypically associated with a prediction that is inconsistent with the derived base rate prediction. In addition to assessing the correspondence between base rates and individual and group predictions, the types of information mentioned during group discussions will be analyzed. The studies will examine the following ideas: 1) The mix of information (descriptive versus base rate) that is commonly known will affect how the prediction task is framed by individuals and subsequently by groups; 2) Task framing will determine how groups reach decisions: analytical frames will be promote information exchange and argumentation (problem solving) and heuristic frames will promote conformity and compromise; and 3) A critical number of members must import an analytical frame into a group for the group to adopt a problem solving approach. Information distributions that are expected to promote analytical frames are ones that result in a critical number of prospective group members having: 1) the necessary information to support an analytical solution and 2) a sense of being better informed than others (thus, promoting more effortful cognitive processing). This work will contribute to our understanding of the processes involved in decision-making and of strategies that could be used to enhance the performance of individuals and groups.

Situations in which individual interests are interdependent with (and partially in conflict with) collective interests are abundant in our daily lives. We encounter social dilemmas in family, work, and social settings, such as when we must decide how much time and effort to contribute to a team project or whether to drive or take public transportation to work. The proposed research examines how subtle variations in the way a message or situation is framed can have important implications for the way people think about and respond to that event. Preliminary research has shown that highlighting the negative collective outcome of non-cooperation can undermine pro-social behavior in dilemmas in which cooperation involves personal restraint from obtaining an immediate benefit, but can promote selfless behavior when cooperation involves giving to the group. The current work explores the affective mediators of this process.

In the proposed study, individuals in three-person groups must decide how to distribute a set of 20 quarters between a personal fund (property of the individual only) and a group fund (shared by everyone in the group). We will measure the impact of three different framing manipulations (decision framing, outcome framing, and payoff structure) on individuals' mood and on their willingness to allocate quarters to their group (cooperation). For the decision framing manipulation, the quarters will either start in the group fund and participants will have the opportunity to take for themselves, or the quarters will start in the personal fund and participants will have the opportunity to give to the group. The quarters in the group fund will face either a bonus or a penalty (outcome frame), and that outcome will be distributed either a little at a time (continuous payoff) or in a lump sum depending on whether or not a specified number of quarters are allocated to the group (step-level payoff).

We predict that the penalty outcome frame will result in a negative mood, which will prompt high levels of giving when the decision involves giving (cooperation) and taking when the decision involves taking (non-cooperation). The pattern of responses will be opposite (low levels of giving and taking, respectively) when the outcome involves a bonus, which is predicted to increase positive mood. Because the step-level payoff provides people with a specific goal upon which to coordinate their behavior, cooperation should be less influenced by mood in the step-level than in the continuous conditions.

Effective coordination is necessary for social interactions as varied as conversation among friends, driving and walking in high-traffic areas, playing sports and working in teams. This project distinguishes between two types of coordination -tacit and explicit coordination while distinguishing matching from mismatching problems. In matching, it is mutually beneficial for actors to choose the same action. Conversely, in mismatching, it is mutually beneficial for actors to choose different actions (e.g., divisions of labor in work teams). Matching and mismatching are expected to foster different interpersonal impression and feelings. These expectations are based on a model that integrates social categorization theory and the concept of focal points from behavioral economics. The proposed work uses a combination of computational modeling and empirical studies. This project will explore 1) how people use social knowledge to coordinate tacitly with others; 2) how successful matching and mismatching emerges over repeated interactions; and 3) how coordination success on matching and mismatching problems affect subsequent interpersonal impressions and team performances. The planned experiments will examine the conditions under which coordination does and does not foster cohesion. Ultimately, the PIs plan to : a) model the effects of task and social factors and decision rules on the development of coordinated matching and mismatching, b) use the modeling results to design efficiently empirical studies with live participants, and c) refine the computational model based on the empirical findings. Findings from the proposed research should contribute to both economic and psychological theorists. The use of a psychologically-themed simulation to represent theoretical assumptions and to derive predictions is both novel and exemplary.

In this project the Principal Investigators will explore how shared meta-knowledge about members' access to information affects participation in collective decisions and how participation rates affect performance on missing links problems. Missing links problems arise when there is a small set of critical information that is known to only one or a few and collective performance depends on these people contributing the critical information. Moreover, others need to ascribe validity to the information and competence to the messenger. The proposed research is guided by the idea that allocating speaking opportunities is a social coordination problem. Participation in group deliberations is characteristically unequal. Some members participate more than others. Existing evidence suggests that the extraverted, dominant, and verbally facile participate frequently but that these traits are not typically associated with knowing more or having expert knowledge. Compounding the problem, high participators are often judged by their peers to be the most competent and knowledgeable. Thus, members who participate frequently are seen as competent which serves to sustain their high rates of participation and enhance their influence. To coordinate participation effectively, members need valid and salient cues to orchestrate the provision and acceptance of speaking opportunities. The investigators will conduct five studies of groups working on missing links problems and reanalyze one set of archival data obtained from a leadership training institute. The explicitness and completeness of meta-knowledge will be varied and the studies will employ several types of tasks: estimation problems, crime mysteries, and quantitative reasoning problems. In these studies, critical information will be given to one or two members; this set of critical information will be necessary to integrate properly the information that all members have. Additionally, in several of the studies, groups will solve several missing link problems with similar structures and receive performance feedback after each problem. The goal is to assess whether participation allocation and collective performance improves with experience of working together in a stable task environment. In addition to collective performance measures, the investigators will record how frequently members participate and what information they contribute. Participants will also rate each others? levels of participation, task-relevant knowledge and competence, and influence after completing the collective tasks. In terms of broader impacts, this research will identify conditions that improve the ability of groups to consider and use critical information in making decisions. Chronicles of failed decision making often identify as a contributing factor failures to communicate and integrate a set of critical information ? in everyday vernacular "failures to connect the dots." For example, the failure to prevent the ?underwear bomber? from boarding a flight from Amsterdam to Detroit on Christmas day, 2009, has been attributed to the failure to disseminate information from the bomber's father and British intelligence within the intelligence community. This in an example of a missing links problem. The proposed studies will identify conditions under which task-relevant knowledge, more than extraversion and social dominance, determine who contributes and who is influential. Because small groups often make important decisions in government and organizations, the proposed work has implications for decision making effectiveness across a wide range of applications from identification of terrorist threats to the prevention of disasters to solving design and technical problems. The proposed research will lay the conceptual and empirical groundwork for designing interventions that reduce the chances of failures to detect missing links in small group decisions.