Libres ou biaisées ? La science des décisions du quotidien | ENS-PSL

Libres ou biaisées ? La science des décisions du quotidien | ENS-PSL

🎙 Constance Destais 👥 75K 📅 July 26, 2026 ⏱ 45 min 👁 592 📄 science communication 🧭 2026-08-03
Available in: English (current) Français

Keywords

biais de confirmationconfianceapprentissage par renforcementdécisions quotidiennesexpérience en laboratoire

Summary

In this conference, Constance Destais, a doctoral researcher at ENS-PSL and PSE, explores the science behind everyday decisions, focusing on cognitive biases, particularly confirmation bias. She defines cognitive biases as systematic deviations from rationality in information processing, giving examples like escalation of commitment and anchoring. The main focus is confirmation bias, the tendency to favor information that confirms existing beliefs, which has been extensively studied in political contexts. The talk extends this to mundane decisions like supermarket choices, using reinforcement learning models to understand how we learn from successes and failures. She explains how these models can be adapted to incorporate confirmation bias, where positive outcomes are weighted more heavily than negative ones. Her own research investigates the role of confidence in modulating this bias, asking whether being more confident makes us more or less biased. She describes experiments where participants choose between symbols with probabilistic rewards, and how their choices reveal learning patterns. The talk suggests that confidence may influence how we integrate feedback, potentially exacerbating or mitigating confirmation bias. Overall, the presentation provides insights into the mechanisms behind our daily choices and the subtle influences that shape them.

190 words

Critical Evaluation

The conference provides a clear and accessible introduction to cognitive biases, particularly confirmation bias, and its relevance to everyday decisions. The speaker, a doctoral researcher, demonstrates a solid understanding of the topic, drawing on established concepts and her own research. The argumentation is logical, starting with definitions, moving to examples, and then presenting experimental paradigms. The use of reinforcement learning models to formalize learning processes is well-explained, making complex ideas understandable. However, the talk lacks explicit citations to specific studies or sources, which would strengthen its scientific rigor. The speaker mentions that confirmation bias has been studied in political contexts but does not name specific researchers or publications. Similarly, her own research is described without detailed methodology or results, limiting the ability to assess its validity. The title accurately reflects the content, and the talk is well-structured. The presence of a brief sponsorship segment (not detailed) does not detract from the content. Overall, the talk is informative and engaging, but a more rigorous citation of sources would enhance its credibility. The speaker’s affiliation with prestigious institutions adds to the reliability, but the lack of concrete references is a notable weakness. The talk successfully bridges theoretical concepts and practical implications, making it valuable for a general audience interested in decision-making and cognitive psychology.

212 words

Title / Content Match

The title accurately reflects the content, which explores cognitive biases in everyday decisions, with a focus on confirmation bias and the role of confidence.

Quality & Reliability

8/10

The talk is given by a doctoral researcher from ENS-PSL and PSE, presenting established concepts in cognitive psychology and her own research. The content is well-structured, references are implicit but credible given the institutional context. No external sources are cited in the description, but the speaker's affiliation and the academic setting support reliability.

Key Moments

Cited Sources

  • Savoirs ENS — Platform hosting the video and other academic content from ENS.
  • ENS-PSL official website — Institutional page of the École normale supérieure - PSL.
  • ENS-PSL LinkedIn — Social media presence of the institution.
  • ENS-PSL YouTube channel — Official YouTube channel of ENS-PSL.

Concurring Sources

Contribution & Novelties

The talk offers a novel perspective by extending the study of confirmation bias from political beliefs to everyday decisions, using computational models. It highlights the potential role of confidence in modulating this bias, which is an emerging area of research. The speaker’s own work contributes to understanding how subjective confidence influences learning and decision-making.

Pour aller plus loin :

104 words

Radar Profile

The radar profile shows high scores in quality of information and reliability, with moderate scores in quantity and technical level. This indicates a well-presented, credible talk that is accessible but not overly technical, suitable for a general audience.

Reliability 8/10