Search

This paper focuses on the interaction of social influence with scientific and intuitive approaches in collective decision-making and under varying degrees of task complexities. By including the effects of being exposed to social information in the form of a peer estimate, we expand on a prior empirical framework by explicitly incorporating the element of social influence in the experimental design. We find that using the scientific approach under social influence results in higher collective performance than intuition. However, due to social influence, the collective accuracy for the intuitive approach improves to a greater degree, especially in the context of highly complex tasks. Regarding the micro-mechanisms that explain these findings, scientific thinkers have been observed to incorporate the provided social information in a manner similar to intuition. Conversely, intuitive thinkers tend to engage in more deliberation and reflection due to having access to social information. Therefore, our empirical results suggest that social influence is a mechanism that enables convergence in outcomes and potentially in terms of the cognitive processes utilized. Through this, not only can social influence further improve the collective accuracy of a group, but it also enables a balance of the limitations and trade-offs that are associated with each prediction approach. [ABSTRACT FROM AUTHOR]

In order to make informed decisions, organizations often rely on aggregating the diverse views, expertise, and opinions of multiple individuals, such as in committees or boards. In this context, the choice of an aggregation structure becomes crucial; it not only significantly impacts the decision outcome but also shapes how individuals perceive and process information during decision-making. Our research closely examines two commonly employed aggregation structures: averaging and voting. Averaging enables the pooling of diverse opinions, providing a holistic viewpoint, but its rigorous approach and cognitive demands may lead to overlooked opportunities. Voting relies on individual expertise and simplifies decision-making, but its binary nature can induce information loss and increased acceptance of less promising opportunities. We analyze the behavioral, cognitive, and statistical elements these structures induce at the individual and organizational levels, focusing on their impact when evaluating uncertain opportunities. Our findings reveal that averaging had slightly lower accuracy at the individual level but higher accuracy at the collective level than voting. Voting led to more investment opportunities and false positives, while averaging resulted in more false negatives. We highlight the trade-offs between voting and averaging when combining multiple individual inputs to extract information signals for quality and to reach organizational outcomes. Our study contributes to understanding information aggregation in organizations and the relationship between artificial and collective intelligence. [ABSTRACT FROM AUTHOR]

Although the effectiveness of the scientific decision-making approach has been demonstrated, most studies have focused on individual entrepreneurs. This panel symposium aims to expand the scope of this research by examining the implications of the scientific approach within large organizations. Our distinguished panelists will present their research, including insights on the challenges and opportunities, decision-making biases, measurement of the approach at scale, and how it differs in application between individual entrepreneurs and large organizations. The panelists will also talk about the topic from a cognitive neuroscience perspective and discuss ways to deal with the problems in current research that focuses on studying the approach and its effects. This panel will provide insights into the effective implementation and navigation of the scientific approach within large organizations while also addressing the unique challenges faced in this context. [ABSTRACT FROM AUTHOR]

Research has demonstrated benefits of the scientific approach to decision-making for individuals' predictive ability in innovation related settings. We move this research from the individual to the collective level. Pursuing this shift in the simplest possible way, we study crowds (i.e., non-interactive groups). We conduct an online experiment and find that crowds that use the scientific approach attain a higher predictive accuracy compared to crowds using the intuitive approach. This advantage increases with the complexity of the prediction task. Further, we elicit the underlying mechanism of this relationship. Importantly, we find that deploying the scientific approach at the collective level is associated with lowering prediction diversity and inducing systematic errors. As a result, the improvements that can be obtained by the scientific approach do not translate completely to the collective level. Theoretical implications are discussed. [ABSTRACT FROM AUTHOR]