Ensaios sobre vieses cognitivos no processo de tomada de decisão gerencial

Abstract

Human behavior and mistakes in judgment have been a subject of study in the field of economics and behavioral finances in order to understand how the individual makes decisions, especially in risky and uncertain environments. In this context, this work had the objective of analyzing, by means of an experiment, if the type (positive or negative) and the level (simple or complex) of economic-financial information influence the cognitive biases of accountants and managers in a managerial decision-making process. The cognitive biases were: Overconfidence, Anchorage Effect and the Confirmation Bias. In addition to bibliometric analyses, the methodology employed was experimental, in which the controlled experimental elaborated for this research was randomly applied by the internet. The treatment factors used were type (positive or negative) and level (simple or complex) of the economic-financial information, and the subjects were company accountants and managers, also considering a control group formed by diverse professionals and students. Thus, this dissertation resulted in six articles, which, in addition to mapping the field of economy and behavioral finances, demonstrated that, in some situations, the managerial information can influence the cognitive biases. The first two articles regarded bibliometric analyses on the research field, showing the importance of the field of economy and behavioral finances, as well as the interaction of the cognitive studied biases with this field of research. The third article brings an experimental proposal to measure the cognitive biases in entrepreneurial decision-making, outlining the conducted experiment. The fourth article aims at analyzing the effects of managerial information over the excess of trust, evidencing that the type of information can influence the excess of trust. The fifth article concerns the analysis of the anchoring effect of accountants and managers and suggests that the type of information influences the anchoring effect in minimum projections related to the low anchor. Finally, the sixth article brings the analysis of the confirmation bias, suggesting that the factors of type and level do not influence managerial decision-making and that the negative information can influence the confirmation bias in the choice of the information most relevant to managerial decision-making. Therefore, this dissertation demonstrates that both accountant and manager as administrators suffer the cognitive biases of excess of trust, anchoring effect and confirmation bias, and that the type and level of information can, in certain situations, influence these biases.