Speech timing variables in children with typical speech

acquisition, speech delay, and suspected apraxia of speech

By J.R. Green, L.D. Shriberg, and T.F. Campbell

Classic descriptions of suspected childhood apraxia of speech include the construct of a “staccato-like” rhythmic quality, yet few acoustic studies have attempted to quantify this percept in children (cf. Shriberg, Aram, & Kwiatkowki, 1997). Related variables in the adult neurogenic literature include the constructs of isochronicity, scanning speech, syllable segregation, and more generally, abnormal speech timing (e.g., Ackerman & Hertrich 1994; Kent, Weismer, Kent, Vorperian, & Duffy, 1999). The present study reports reference data for the distributions and sequential occurrences of pauses and speech events in the conversational speech of three groups of 3 to 6 year-old children. Using several auditory-perceptual speech metrics and prosody screening methods (Shriberg, Austin, McSweeny, Lewis, & Wilson, 1997a; 1997b) 30 children were classified as having normal speech acquisition, 30 as having moderate to severe speech delay, and 12 were classified as having suspected apraxia of speech. Matlab routines were developed to characterize the distributional characteristics and regularities in the serial ordering of pause and speech events sampled in 24 utterances from each speaker. The focus of the presentation is on the contributions of findings toward development of an acoustic marker of childhood AOS.