Key points are not available for this paper at this time.
A consensus transcription procedure is presented, and representative data are reported. The procedure includes explicit response definitions for mapping narrow phonetic transcription onto dichotomous, nonerror/error scores and 17 rules for obtaining consensus between two transcribers. The data include descriptive statistics for consensus rule use based on over 18,000 sounds transcribed during a normative study of speech sound acquisition in children aged 3–6 years. Validity and reliability issues in phonetic transcription are discussed.
Shriberg et al. (Sat,) studied this question.