Perceptual and acoustic reliability estimates for the Speech Disorders Classification System (SDCS).

Area of Science:

• Speech-Language Pathology
• Linguistics
• Audiology

Background:

• The Speech Disorders Classification System (SDCS) is a classification system for paediatric speech sound disorders.
• The SDCS utilizes perceptual and acoustic data reduction methods to analyze speech, prosody, and voice.
• Reliability estimates are crucial for validating classification systems in speech sound disorder assessment.

Purpose of the Study:

• To provide reliability estimates for the perceptual and acoustic methods within the SDCS.
• To assess the intra-judge and inter-judge agreement for SDCS methods.
• To support the classification of speech competence, precision, and stability in speakers.

Main Methods:

• Re-measurement of speech samples from 10 speakers (5 with motor speech disorder, 5 typical).
• Utilized narrow phonetic transcription and prosody-voice coding for perceptual analysis.
• Employed acoustic analysis including phoneme segmentation, fundamental frequency tracking, and formant estimation.

Main Results:

• Perceptual methods achieved point-to-point percentage of agreement as high or higher than previous studies.
• Acoustic analysis tasks showed agreement ranging from the mid-70% to 100%, with most in the mid-80% to mid-90% range.
• High reliability was found for both perceptual and acoustic measurement methods.

Conclusions:

• The findings support the reliability and validity of the perceptual and acoustic methods used in the SDCS.
• The SDCS is a dependable tool for describing and classifying speakers with speech sound disorders.
• This research contributes to the robust assessment of paediatric speech sound disorders.