Normative data and physical measurements for a nonsense CVC intelligibility test
Christine M. Rankovic, Dianne J. Van Tasell · The Journal of the Acoustical Society of America · 1988
A nonsense CVC stimulus set has been developed for investigations into the application of the articulation index (AI) to speech recognition testing with hearing-impaired persons. Each test list consists of 57 items drawn at random from a corpus of 228 items comprising syllables spoken by four talkers. Crossover functions support the similarity of this stimulus set to that of French and Steinberg [J. Acoust. Soc. Am. 19, 90 (1947)] in terms of frequency importance function. Because the number of vowels was restricted to three, the long-term spectrum of the stimulus set is less smooth than that usually associated with connected speech. The AI versus percent-correct phoneme score functions have been obtained from normally hearing subjects. Comparisons with published data for similar stimulus sets will be made. Talker differences, reliability of subject performance, and preliminary data from hearing-impaired subjects will be reported. [Work supported by NICHHD T32 HD-07151 and NINCDS NS 12125.]