Main content
Adjudicating between auditory models using individual variation in infant vowel discrimination
Date created: | Last Updated:
: DOI | ARK
Creating DOI. Please wait...
Category: Project
Description: Many different models have been proposed to capture acoustic properties of speech - but which describes infant perception? Discrimination scores were gathered from individual infants for two vowel contrasts. These and other categories were subsequently elicited from each infant's primary caregiver. Regression analyses were used to assess fit between individual variation in infant vowel discrimination, and auditory contrast estimates for the relevant and control contrasts in caregiver speech. The latter were calculated using three models of speech audition; formants, RASTA-PLP and CASP. None of these models were significantly related to infant discrimination scores. As a follow-up, another independent set of infant discrimination scores and caregiver productions was gathered; however, due to a failure to replicate the previous discrimination results no further analyses were conducted.