Schematic pitch coding: a new, more efficient method for measurement of infant-directed speech

Infant-directed speech is an important but difficult-to-measure aspect of caregiver-infant interaction. Specialized training in speech science is required and processing is computationally intensive and time consuming. We developed a computer-assisted method of measuring infant-directed speech that requires little specialized training, has high concurrent validity with alternative approaches, and has improved efficiency in estimating frequency and temporal parameters. In Schematic Pitch Coding, an operator uses a computer interface to play digitized speech samples, view the corresponding narrow-band spectrograms, and mark points of inflection in fundamental frequency (f) using a computer mouse. Custom software generates graphic representations of f contours and quantitative estimates of the mean and standard deviation of f, utterance duration, and turn-taking pauses. Our current system, which is not optimized for speed, can process a 30-second sample of infant-directed speech in approximately a half-hour, which makes feasible its application in large-scale studies of infant-directed speech.