Notched-noise measures of frequency resolution in children: What acoustic cues are available at children’s threshold levels?

Abstract

It has been suggested that young children’s masked detection thresholds are obtained at signal-to-noise ratios for which the signals produce changes in multiple features of the masker [Allen et al., J. Acoust. Soc. Am. (in press)]. In this study 500-, 2000-, and 4000-Hz tonal signals were added to flat- and notched-spectrum noise maskers at signal-to-noise ratios corresponding to children’s threshold levels measured in a previous study of frequency resolution [Allen et al., J. Speech Hear. Res. 32, 317–324 (1989)]. The magnitudes of level, frequency, and periodicity cues were determined, assuming several degrees of frequency selective filtering. Results suggested that for all signal plus masker combinations, independent of frequency and masker spectrum, children’s thresholds had been obtained at signal-to-noise ratios for which little or no frequency selectivity would have been required to produce potentially discriminable level differences [Jensen and Neff, Psychol. Sci. 4, 1-4-107] of approximately 6 dB, which were often accompanied by average frequency and/or periodicity differences. Age-related changes in frequency selectivity therefore may not be required to explain children’s higher masked detection thresholds as they are obtained at signal-to-noise ratios where multiple cues are present. [Work supported by NSERC.]