A Imparcialidade Opaca: testando empiricamente a (in)existência de vieses cognitivos nos acórdãos do Tribunal Regional Federal da 5ª Região.

• Ano de defesa: 2025
• Autor(a) principal: Lustosa, João Paulo Pessôa Pereira
• Orientador(a): Gouveia, Lucio Grassi de
• Banca de defesa: Gomes Neto, José Mário Wanderley, Costa, Eduardo José da Fonseca, Abreu, Rogério Roberto Gonçalves de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Católica de Pernambuco
• Programa de Pós-Graduação: Mestrado em Direito
• Departamento: Departamento de Pós-Graduação
• País: Brasil
• Palavras-chave em Português: Dissertações; Juízes - Decisões; Julgamento; Pesquisa quantitativa; Brasil - Tribunal Regional Federal da 5ª região
• Palavras-chave em Inglês: Dissertations; Judges - Decisions; Judgment; Quantitative research; Brazil - Federal Regional Court of the 5th region
• Área do conhecimento CNPq: CIENCIAS SOCIAIS APLICADAS::DIREITO
• Link de acesso: http://tede2.unicap.br:8080/handle/tede/2019

What cognitive biases can influence the judges of the Federal Regional Court of the 5th Region (TRF5) in decision-making processes? This research seeks to empirically test the (non)existence of cognitive bias in the judgment of civil appeals, judged by the panels of TRF5 in 2023, hypothesizing that the judges of the researched court are likely influenced by anchoring, status quo, and group biases in their cognitive decision-making process. The study's relevance arises from potential influence of heuristic errors on the judicial activity, which could obscure the impartiality of the judges from the court responsible for federal justice in six Northeastern states (AL, CE, PB, PE, RN, and SE). These impacts could be mitigated or better understood based on the results of this investigation, also allowing for application in other courts across the country. The methodology employed is empirical-quantitative, descriptive, and exploratory. Initially, a literature review on civil appeals and the jurisdiction of TRF5 panels will be conducted, followed by an in-depth study on cognitive bias. Subsequently, a dataset of rulings issued in civil appeals judged by TRF5 panels in 2023 will be collected, classified, and statistically analyzed using logistic regression (logit) to examine the correlation between different variables, aiming to diagnose the interplay of influences among them. In the end, secondary hypothesis 1 (the Reporting Judge tends to uphold the appealed decision) presents a statistically significant constant; secondary hypothesis 3 (judgments tend to favor the public entity) was confirmed; whereas secondary hypothesis 5 (appeals in cases involving high claim amounts tend to be denied) was statistically rejected, paving the way for the development of de-biasing algorithms.