R. Scott Tindale - US grants

Committees, focus groups, and self managed work teams have become permanent fixtures in many organizations. Reliance upon small groups in our society can be attributed in part to the pervasive belief that groups outperform individuals working alone. The power extended to small, decision making groups is especially interesting when viewed in light of the research exploring group and team efficacy. An extensive research literature supports the notion that group performance is frequently superior to individual performance, but recent findings suggest that groups can also increase the likelihood of error under certain circumstances. In our program of research we have demonstrated that for some decision making tasks, incorrect minorities can sway correct majorities toward their preferred incorrect alternative (e.g., Tindale, Smith, Thomas, Filkins, Sheffey, 1996.) In using the notion of a shared task representation, we have been able to predict when groups will perform better or worse than individuals working alone. With the proposed research we will attempt to generalize our findings to new domains, define the origins of shared representations (i.e., culture, ideology) and how these representations operate in the various contexts in which groups make their decisions. Our ultimate aim is to devise ways with which we can reduce the potential biasing effects of shared representations.

Small decision-making groups resolve many of the most important issues/questions in organizations and society. Such groups bring more resources than any single individual to the task at hand. There is also a fair amount of research support for the superior decision-making capabilities of groups as compared to individuals. However, recent work has shown that the same processes that make groups superior decision-makers and problem-solvers can under certain circumstances, lead them astray.

Groups perform better than individuals in many situations because when a correct or optimal alternative is proposed, groups have the ability to recognize the "correctness" of the alternative. Our recent work has shown that a group's ability to do this depends on an appropriate task representation shared by the group members. In most situations, the group members share an appropriate representation. However, we have shown that in situations where group members share a faulty task representation, group often perform worse than individuals. The research contained in the current proposal will attempt to further our understanding of how such shared task representations affect group decision-making. In addition, the research will address how competing shared representations - specifically representations at different levels of generality - affect group decision outcomes and processes. The overarching goals of the research are show how to enable groups to use appropriate task representations when they exist, but to learn how to prevent them from using faulty representations that are initially shared among the group members.

People are often asked to make decisions about issues that are complex and involve uncertainty in terms potential outcomes. One of the ways people attempt to insure that they make the best decision possible is to discuss such decisions with others. Many formal decision making settings require groups (e.g., juries, corporate boards, etc) to make important decisions to insure that multiple people have thought about and reached consensus on the issue at hand. However, recent research has shown that under certain circumstances, having groups, as opposed to individuals, make decisions can lead to more biased or less accurate decisions and judgments. The explanation for these findings is that the group members share an inaccurate or inappropriate perspective on the particular task. Thus, the group members tend to provide support for each other on incorrect decisions rather than providing a corrective influence. The research proposed here will attempt to further understand how such shared misrepresentations operate in groups and will also explore conditions that may correct such misrepresentations. By manipulating the number of group members who share the misrepresentation and by providing training to some group members to correct the misrepresentation, the investigators hope to discover the conditions under which group or individual decision makers can overcome decision biases and use information provided by others to make better decisions.

Much of the decision-making and problem-solving work done in organizations involves groups or teams. As a consequence, a central component of high-level performance involves optimal information processing by groups. Although research on group information processing has increased substantially over the past 10 to 15 years, a number of key questions remain unanswered. The current proposal attempts to directly address one of these key questions: what is the optimal way to distribute information among group members. Work on group memory has shown that group information storage capacity is maximized when each member is responsible for different types of areas of information. However, research on group information processing has shown that information use is maximized when information is shared or known by many if not all group members. The implications of these two findings are diametrically opposed. Thus, the purpose of the proposed research is to systematically investigate group memory and performance across different levels of information redundancy across members. Using a number of different task domains, Dr. Scott Tindale of Loyola University of Chicago and Dr. Amanda Dykema-Engblade of Northeastern Illinois University will investigate how best to distribute information among group members so as to optimize both storage capacity and use. They will also assess the effects of shared mental models on group memory and decision performance. Two different types of mental models will be addressed: models of the group in terms of roles and expertise, and models of the task in terms of key task components and how those components must be integrated for successful performance. The researchers predict that when group members share accurate mental models of both types, they will be able to handle greater amounts of information and put it to better use than they would be able to do under conditions where such mental models are lacking. Overall it is expected that some level of information redundancy will be necessary for optimal group functioning, but that shared mental models will reduce the amount of redundancy required. The results of this research will help organizations to design group task environments for maximum performance.

This research examines the conditions that prompt groups to underutilize advice, and asks whether groups might sometimes underutilize advice to a lesser extent than individuals do. Because both individual and group decision makers often have advice available to them, gaining a better understanding of when they will be most likely to take advantage of that advice may help us understand why certain individuals or groups seem to make good decisions and why and when they might be prone to make bad ones. This research will contribute to national health and to US competitiveness to the extent it informs the design of processes that promote effective use of advice in individual and group decision making so as to improve individual outcomes and organizational performance.

Decades of research have shown that decisions made in a social context, where many people's opinions are integrated before a final judgment is reached, tend to be better than decisions made by lone individuals (Kerr & Tindale, 2004; Larrick & Soll, 2006; Larson, 2010). Unfortunately, most people do not realize the benefits of integrating advice from others into their own judgments (Bonaccio & Dalal, 2006), and so they either underutilize advice or ignore it altogether. Further, it is not just individual decision makers that underutilize advice; decision-making groups do so as well. Indeed, one recent study found that small decision-making groups underutilized external advice to an even greater extent than individuals do, and, like individuals, they paid a price in accuracy for doing so (Minson and Mueller, 2012). The present research focuses on the conditions that prompt groups to underutilize advice, and asks whether groups might sometimes underutilize advice to a lesser extent than individuals do. Group members are typically very confident in their decisions right after they have reached a consensus, but beforehand, and especially when there is substantial diversity in their members' individual decision preferences, they are apt to have relatively low confidence in their judgments (Tindale, 1989). Because confidence likely plays an important role in determining openness to advice, we predict that groups will make much better use of advice prior to consensus when their members initially hold diverse decision preferences. In our first study, advice will be provided to both individuals and groups at various stages of the decision making process in order to learn more about when, during that process, groups and individuals are most likely to make good use of advice. Our second study will then explore these same processes in a more ecologically valid situation, where decision-makers must actively seek-out advice, rather than the advice being provided to them automatically.