Defense in english
Neural and cognitive bases of confirmation bias-induced interference in declarative memory performance
Confirmation bias is a well-described and ubiquitous cognitive behavior whereby novel information from the environment is over-valued when it confirms and under-valued when it disconfirms previously consolidated cognitive content (e.g. beliefs, learned associations, etc.). The maladaptive responses this phenomenon can give rise to are implicated in social problems such as the spread of “fake news” and vary according to both contextual complexity and the mental state of the subject.
Nevertheless, very little research has been dedicated to understanding the neural mechanisms or evolution underpinning this spontaneous human cognitive response to novel information. Thus, we designed a mouse model for confirmation bias-like behavior, enabling exploration of its cognitive and neurobiological underpinnings and their evolution. Our model is based on a cognitive level definition of the phenomenon; over-valuation of novel environmental elements which confirm and under-valuation of novel environmental elements which disconfirm a previously consolidated cognitive content. Our results to this point (using a two-task, two-context radial maze protocol) show a strong bias effect which is observable as a deviation in the performance of a classical declarative memory task, the persistence of which is trial-complexity dependent.
Detailed behavioral analysis has enabled us to identify several more basic cognitive components impacting the bias effect, such as adaptive forgetting and the exploration/exploitation balance. These cognitive components have been identified with specific neural circuits whose activity is susceptible to intervention and/or monitoring in freely moving task-performing animals. They are also implicated in many psychiatric conditions (depression, schizophrenia, etc.) making of this model a novel tool for pre-clinical research of which we are developing a human version for clinical research and a computational version for formulating and testing predictions.
|Mme. RAVEL Nadine, D.R.||Université de Lyon, France||Présidente|
|M. COUTUREAU Etienne, D.R.||Université de Bordeaux, France||Examinateur|
|M. MALLERET Gaël, C.R.||Université de Lyon, France||Rapporteur|
|M. PALMINTERI Stefano, C.R.||École Normale Supérieure, Paris, France||Rapporteur|
|Mme. MARIGHETTO Aline, D.R.||Université de Bordeaux, France||Directrice de thèse|
|M. MARSICANO Giovanni, D.R.||Université de Bordeaux, France||Invité|