Technical Program

Paper Detail

Paper: PS-1A.36
Session: Poster Session 1A
Location: H Lichthof
Session Time: Saturday, September 14, 16:30 - 19:30
Presentation Time:Saturday, September 14, 16:30 - 19:30
Presentation: Poster
Publication: 2019 Conference on Cognitive Computational Neuroscience, 13-16 September 2019, Berlin, Germany
Paper Title: The origin of fixed, history-independent choice biases of rodents in perceptual decision making tasks
Manuscript:  Click here to view manuscript
License: Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
Authors: Gabor Lengyel, Central European University, Hungary; Alexandre Hyafil, Universitat Pompeu Fabra, Spain; Jozsef Fiser, Central European University, Hungary; Jaime de la Rocha, Institut d’Investigacions Biomediques August Pi i Sunyer, Spain
Abstract: Perceptual decision making is typically described with classical models (Signal Detection, Bayesian Decision, and Drift Diffusion Models) that distinguish modules for sensory processing, decision making and additional post-decisional processes with separate bias terms for each module. At the behavioral level, animal and human decision making 2-AFC studies provide ample evidence for choice-biases during such tasks. A subset of these biases are history-dependent and can be directly linked to the rewards, responses and stimuli, but the origin of the other subset of fixed history-independent biases in these tasks is still largely unknown. Here, we investigated whether these fixed biases could originate from the decision module as defined by the classical models or they must be related to post-decisional processes. We designed an interaural amplitude discrimination task with rodents, with the amplitude difference and the maximum intensity level modulated simultaneously, and found that the performance of all rats decreased asymmetrically at the two response sides as a function of decreasing intensity level. Through computational analyses, we show that these fixed biases cannot be explained within the decision module and are only compatible with post-decisional biases.