Development of formant frequency distributions in American English-speaking elementary school-aged children: A longitudinal study

Abstract

Speech development enters a critical phase in the early elementary years, when learned patterns of phonology and motor control become increasingly entrenched, and developmental speech-language disorders become more difficult to treat. Much attention has been given to early speech development in the preschool years, and a few studies have examined the acoustics of speech development cross-sectionally from Kindergarten to adulthood (see Kent and Vorperian, J Commun Disorders 74, 74–97 (2018)]. To the best of our knowledge, there are no longitudinal acoustic studies of the speech of typically developing children between 5 years and puberty. Although cross-sectional studies provide a general picture of growth, they cannot reveal specific patterns of development and variability. We present acoustic data from an annual, longitudinal study of 8 children between 6 and 10 years of age. All participants are native speakers of Midwest American English. Stimuli included words and sentences from the Goldman-Fristoe Test of Articulation 3rd Edition, as well as a spontaneous speech sample and nonce syllables embedded in a carrier phrase. Frequencies for the first four formants were obtained across the entire stimulus set with Praat, and used to generate formant frequency distributions for each child, in each year.