A procedure to select and train persons for narrow phonetic transcription by consensus

Abstract

This report describes a study of procedures to select and train persons for research positions as phonetic transcribers. A two-hour, job-sample procedure included repeated trials on transcription tasks with interim feedback on performance. In addition to the quantitative data on transcription potential, the procedure allowed examiners to observe and rate subjects' cognitive and affective responses during simulated training conditions. A total of eight subjects were tested and two transcribers were subsequently selected and trained for narrow phonetic transcription by consensus (Shriberg, Kwiatkowski and Hoffmann, 1984). Discussion of the quantitative data and anecdotal observations during both phases of the study focuses on applied implications of a three-factor view of the competent transcriber.