Since the early days of psychology, practitioners have recognised that metacognition - or the act of thinking about one's own thinking - is intertwined with our experience of the world. In the last decade, scientists have started to understand metacognitive signals, like judgments of confidence, as precise mathematical constructs. Confidence can be conceived of as an internal estimate of the probability of being correct. As such, confidence influences both advice seeking and advice taking while allowing people to optimally combine their views for joint action and group coordination. This work begins by exploring the idea that confidence judgments are important for monitoring not only uncertainty associated with one's performance but also, thanks to their positive covariation with accuracy, the reliability of social advisers, particularly when objective criteria are not available. I present data showing that, when adviser and advisee's judgments are independent, people are able to detect subtle variations in advice information, irrespective of feedback presence. I also show that, when such independence is broken, the use of subjective confidence to track others' reliability leads to systematic deviations. I then proceed to explore the differences existing between static and dynamic social information exchange. Traditionally, social and organisational psychology have investigated one-step unidirectional information systems, but many real-life interactions happen on a continuous time-scale, where social exchanges are recursive and dynamic. I present results indicating that the dynamics of social information exchange (recursive vs. one-step) affect individual opinions over and above the information that is communicated. Overall, my results suggest a bidirectional involvement of confidence in social inference and information exchange, and highlight the limits of the mechanisms underlying it.

Les régions construisent leur avantage économique en s'appuyant sur leur capacité à mobiliser et à canaliser les connaissances. L'enjeu de cette recherche est de comprendre la dynamique à l’œuvre dans les relations inter-acteurs qui explique la sous-performance des uns ou les capacités créatives des autres par le biais de la diffusion de connaissances ou de l'apprentissage. L'enjeu est de cerner la manière dont les individus, les firmes et les institutions interagissent ainsi que les effets de leur interaction sur la performance régionale en matière d'innovation. Nous proposons d'approcher ce sujet d'étude sous l'angle de la théorie des conventions, en explorant d'autres dimensions importantes de la créativité et de l'apprentissage, particulièrement en articulant les notions de confiance et de croyance. Les individus agissent et interagissent dans un contexte social régi par des conventions. Une action ne peut s'effectuer sans l'existence d'un certain nombre de conventions. Un des problèmes majeurs des conventions est qu'elles limitent souvent l'ensemble d'actions possibles. Elles structurent et contraignent l'action. Elles ne sont pas sans lien avec les notions de confiance et de croyance en question. / Regions build their economic advantage in relying on their ability to mobilize and channel knowledge. The aim of this research is to understand the dynamics at work in the inter-actor relationship which explains the underperformance of some actors or the creative abilities of others through dissemination of knowledge or learning. The challenge is to identify how individuals, firm sand institutions interact and the effects of their interaction on regional performance in terms of innovation. We propose to approach this subject of study under angle of conventions theory, exploring other important dimensions of creativity and learning, particularly by articulating the notions of trust and belief. lndividuals act and interact in a social context governed by conventions. An action can not take place without the existence of some number of conventions. One major problem of conventions is that they often limit the set of possible actions. They structure and constrain action. They are not unrelated to the notions of trust and belief in question.