Bhandari, P. 1 & Soto, D. 1, 2
1 Basque Center on Cognition, Brain and Language
2 Ikerbasque, Basque Foundation for Science
The predictive processing framework posits that we continuously make predictions of occurrence of an event based on the prior probability of such occurrences. Studies have shown that in low-level visual perception tasks when predictions do not match the stimuli (i.e. prediction error), participants? metacognitive sensitivity (i.e. how confidence judgments relate to task accuracy) is reduced compared to when the stimuli match with the prediction. However, it is not known whether such relationship extends beyond low-level perceptual tasks to higher level cognitive processes. In this study, we use object categorization task (living / non-living judgment) which could be presented in different languages. 20 Spanish-Basque bilingual participants are presented with Spanish and Basque words. Participants are required to perform two-alternative forced choice (2AFC) task to respond whether the word presented was a living being or a non-living being, immediately followed by a confidence rating on a scale of 1 to 4. Staircase method is used to manipulate the visibility of the words and maintain the participants? accuracy at 70%. The predictability is manipulated by presenting Spanish words in 80% of the trials, i.e. predictable trials and Basque words in 20% of the trials, i.e. non-predictable trials. We hypothesize is that metacognitive confidence will be influenced by language predictability, while there may not be a significant effect of language predictability on first-object performance categorization task. The results presented will be discussed in connection with existing models of metacognition and predictive processing.