In our daily lives, we interact in repeated occasions with the same perceptual information. Often, each of these interactions lead to different decisional outcomes between one occasion to another. Perceptual decision-making can benefit from the information that we gathered previous times that we encountered the same situation. In this project, we wanted to study which mechanisms determine performance of participants in a repeated perceptual decision-making task. For instance, participants might integrate efficiently new information with the previous one (i.e. maximum likelihood estimation), be more sensitive to confirmatory evidence, while reducing sensitivity to disconfirmatory evidence (i.e. confirmation bias, Talluri et al,. 2018) or simply repeat the same response regardless of the new collected information (i.e. decision bias). To test these hypotheses, we designed an experiment where participants had to estimate the mean orientation of 6 sequentially presented Gabor patches. We found that participants’ responses improved after repeated exposure to the same stimuli sequence. Interestingly, this enhancement was primarily achieved through a “disconfirmatory mechanism” by which participants weighted more sensory evidence that disconfirmed previous choices on the same stimulus sequence. Our results demonstrate that humans, when faced with exactly the same information in repeated occasions distribute their perceptual resources asymmetrically, overweighting incongruent information with their previous believes.
| Date of Award | 2019 |
|---|
| Original language | English |
|---|
| Awarding Institution | - Universitat de Barcelona (UB)
|
|---|
| Supervisor | Alexis Perez Bellido (Director) |
|---|
Disconfirmation bias explains perceptual enhancement in repeated sensory experiences
Bujalance, A. (Author). 2019
Student thesis: Dissertation (TFM)
Bujalance, A. (Author),
Perez Bellido, A. (Director),
2019Student thesis: Dissertation (TFM)
Student thesis: Dissertation (TFM)