Human Compliance with Normative Principles in Argumentation: Effects of Naturalness Bias and Graphical Representation - Archive ouverte HAL
[go: up one dir, main page]

Communication Dans Un Congrès Année : 2024
Human Compliance with Normative Principles in Argumentation: Effects of Naturalness Bias and Graphical Representation
"> University of Belgrade, Faculty of Philosophy, LIRA Lab
2 SyCoSMA - Systèmes Cognitifs et Systèmes Multi-Agents (France)
"> SyCoSMA - Systèmes Cognitifs et Systèmes Multi-Agents
3 CRIL - Centre de Recherche en Informatique de Lens (Faculté Jean Perrin Rue Jean Souvraz - SP 18 - 62307 LENS CEDEX - France)
"> CRIL - Centre de Recherche en Informatique de Lens
4 CNRS - Centre National de la Recherche Scientifique (France)
"> CNRS - Centre National de la Recherche Scientifique

Résumé

Argumentation theory examines how conclusions are derived or refuted through logical reasoning, playing a crucial role in human interaction and decision-making. In artificial intelligence, computational argumentation leverages formal models to aid in decision-making processes. This paper investigates the influence of argument content (specifically the naturalness bias) and graphical representation on participants' adherence to the simple principles of reinstatement and void precedence principles. We conducted experiments testing three hypotheses related to participants' rationality in evaluating arguments with and without graphical aids and bias-provoking content. Contrary to our expectations, neither the presence of graphical representations nor the type of content significantly impacted participants' compliance with the reinstatement principle. Additionally, the graph did not enhance understanding, suggesting the need for instructional aids. Our findings challenge previous studies and highlight the complexity of factors influencing argument evaluation.

Fichier principal
Vignette du fichier
argxaii.pdf (1.25 Mo) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-04720157 , version 1 (03-10-2024)

Licence

Identifiants
  • HAL Id : hal-04720157 , version 1

Citer

Marija Petrović, Predrag Teovanović, Danka Purić, Bruno Yun, Caren Al Anaissy, et al.. Human Compliance with Normative Principles in Argumentation: Effects of Naturalness Bias and Graphical Representation. The 2nd International Workshop on Argumentation for eXplainable AI (ArgXAI 2024), Timotheus Kampik; Kristijonas Čyras; Antonio Rago; Oana Cocarascu, Sep 2024, Hagen, Germany. ⟨hal-04720157⟩
23 Consultations
32 Téléchargements

Partager

More