Correlations between auditory capabilities and measures of phoneme perception
Blas Espinoza-Varas, C. S. Watson, D. A. Geddes · The Journal of the Acoustical Society of America · 1986
Performance of 32 normal-hearing listeners on the test of basic auditory capabilities [Watson et al., J. Acoust. Soc. Am. Suppl. I 71, S73 (1982)] was correlated with each of the following measures of phoneme perception (obtained with the CUNY Nonsense Syllable Test): (l) hit probability for the overall test, and for each individual phoneme; (2) P(C)max for identification of each individual phoneme; and (3) P(C)max for discrimination of each possible phoneme pair within a set of phonemes. Relative to the total number of correlations that can be calculated, the percentage of significant correlations (p < 0.05, n = 32) were 0% for overall hit probability, 10% for phoneme-specific hit probability, 17% for identification P(C)max and 10% for pairwise discrimination P(C)max. These results support the hypothesis [Espinoza-Varas and Watson, J. Acoust. Son. Am. Suppl. 1 78, S47 (1985)] that auditory capabilities may be correlated more strongly with the ability to process specific phonemes than with overall measures of speech processing (e.g., whole-test speech discrimination scores). It was also found that a number of phonemes yielding either near-chance or near-perfect performance contribute little or nothing to the correlations or, presumably, to the information provided by the overall speech score. [Work supported by NIH and AFOSR.]