Grace Brown awarded NSF GFRP

Modeling perception in Autism.

Big congratulations to Baggett Fellow Grace Brown, whom the NSF has awarded a Graduate Research Fellowship, for her proposed project on "Constructing a probabilistic model of speech and sensory awareness in Autism," which is abstracted below. 

Constructing a Probabilistic Model of Speech & Sensory Perception is Autism 

Atypical sensory perception is one of the most common symptoms of autism, and yet we do not have a working cognitive model of sensory perception in autism. Sensory perception is a building block of higher-order cognitive representations, meaning that changes in sensory perception have the potential to impact domains like social communication or speech perception. 

I aim to construct a model of sensory perception that explains the differences between neurotypical and autistic perception and that simultaneously accounts for the individual variability of perceptual abilities among autistic people. This model will connect theories of neural encoding and probabilistic models of computation in perception. Specifically, I propose that Bayesian models of computation are compatible with a neural account of autism that attributes sensory patterns like hypersensitivity to an excitatory/inhibitory imbalance within neural populations involved in sensory processing. Using speech perception as a case study, I will compare this model to my own experimental data that will create a profile of the speech perception abilities of autistic people relative to their typically developing (TD) peers. The overarching goal of this work is twofold: to construct a comprehensive model of speech perception in autism that can be generalized to other sensory modalities and to improve behavioral accounts of speech perception in autism by producing highly controlled experimental data.