uncertainty and prediction

uncertainty

refers to situations where the outcome, cause or meaning of something is unclear

event uncertainty

not knowing what will happen

causal uncertainty

not knowing what causes something

perceptual uncertainty

not sure what we are seeing

social uncertainty

not knowing why someone behaved in a certain way

normative model

describe the most ideal way we make decisions and resolve uncertainty even if we don’t always use it

bayesian inference

combines prior beliefs with new evidence to update posterior belief - basically we update predictions with new data

implications of bayesian inference for cognition

suggests that the brain uses probability and predictions help us to detail with uncertainty

but

evolution favours good enough solutions not perfection

errors such as base rate fallacy

bayesian inference in multisensory integration

the brain combines multiple sources of information through integration strategies

simple average

equal weighting of senses

weighted average- precision weighting

more reliable senses receive greater weight

how do we process multiple sources

the brain integrates information by weighting sources according to their reliability

prediction in social cognition

social understanding involves combing prior expectations and new sensory input from observing actions

precision weighting in social perception

people use predictions about efficient actions and when sensory information is uncertain- predictions influence perception

predictions about goals and intentions

when interpreting actions- people integrate prediction and sensory evidence