A carregar...
Projeto de investigação
The Social Amplification of Bias
Financiador
Autores
Publicações
The social amplification of bias
Publication . Mendonça, Cristina; Mata, André; Ferreira, Mário Augusto de Carvalho Boto; Fiedler, Klaus
The area of judgment and decision-making became very influential under the heuristics-andbiases research program by revealing that the prevailing economic theories lacked realism: The human being is not, as these theories had assumed, perfectly rational. Instead, people frequently rely on simplified, mostly automatic and unconscious strategies (i.e., heuristics) that enable the human mind, with its limitations, to provide acceptable judgments and decisions, but that can also lead to systematic errors (i.e., biases) under certain conditions.
Yet, the judgment and decision-making research itself still neglects essential aspects of reality that can have an important influence on the way people judge and make decisions.
Among these aspects is the fact that, as people live in societies, they receive information from, and transmit it to, other people. The main hypothesis explored in the current dissertation is that this social dynamic will lead to a social amplification of bias: As information travels from one person to the next, the message will aggregate individual biases leading to messages that are progressively more biased the further they travel from their source. In four experimental chapters, the social amplification of bias hypothesis was tested using the serial reproduction paradigm. In this paradigm, communication chains are formed using the responses of one participant (e.g., the recall of a text) as the stimuli to be presented to the next participant, thus recreating the social dynamic of receiving and transmitting information in the laboratory. These studies supported the social amplification of bias hypothesis, and did so covering different judgment and decision-making domains (risk perception, illusory correlations, denominator neglect, and cognitive reflection), using different types of response formats (frequency estimates, forced recognition, free recall), andincluding samples from Europe and the US, online and in the lab. The dissertation ends by discussing implications, future research, and potential modelling and debiasing techniques.
Real people or mere numbers? The influence of kill-save ratios and identifiability on moral judgements
Publication . Costa-Lopes, Rui; Mata, André; Mendonça, Cristina
In moral dilemmas, decision-making can be based on more utilitarian or deontological reasoning. In two experimental studies, we manipulated the number of people to be sacrificed (1 to save 5 vs. 3 to save 5) and whether personalizing information about them was presented. Results provide the first evidence of how the effects of kill-save ratios and identifiability of the potential victims are contingent on one another. Specifically, this research shows that when individuating information about the potential victims is present in a trolley dilemma, participants are more reluctant to sacrifice three persons to save five than to sacrifice one person to save five. When such individuating information is not present, the acceptability of sacrificing the victims does not depend on their number.
Unidades organizacionais
Descrição
Palavras-chave
Contribuidores
Financiadores
Entidade financiadora
Fundação para a Ciência e a Tecnologia
Programa de financiamento
Número da atribuição
PD/BD/113491/2015
