How perceptual inference changes with age: Behavioural and brain dynamics of speech perception

Our perceptual interpretation of noisy and uncertain environmental signals does not only rely on their bottom-up sensory analysis but is also critically shaped by our acquired knowledge about their underlying causes. To compensate for declining sensory fidelity, older adults tend to rely more strongly on prior knowledge, biasing their perception towards expected sensory events. While this strategy may be generally beneficial, it can turn maladaptive when it is applied inflexibly. However, in how far aging impairs the flexible adaptation of perceptual strategies in uncertain environments remains insufficiently understood. Moreover, we have limited understanding of how adverse effects of age on perceptual accuracy are linked to the computational and neural level. Using the auditory modality and the well-known lexical bias effect on speech perception as a model system, we will investigate how aging affects the flexibility with which prior knowledge and sensory information are integrated. In cross-sectional samples of healthy younger and older adults, the reliability of sensory information and long-term lexical knowledge will be systematically altered to test the hypothesis that older age leads to the application of increasingly precise perceptual priors and reduced belief updating. The project features an extensive behavioural and psychophysical protocol that is combined with an established mathematical model of Bayesian perceptual inference and model-based functional brain imaging to provide an multi-level account of age-related changes in perceptual inference.