Hearing in the speech range as a key to a puzzle of loudness constancy and loudness adaptation

Abstract

Loudness adaptation has been a controversial phenomenon, partly due to differences in subject behavior at different frequencies, and partly to conflicting results from studies using different methodology. In this study, parallel experiments were run contrasting two methodological paradigms, the Ipsilateral Comparison Paradigm (ICP) and Simple Adaptation (SA). Both used magnitude estimation to assess loudness adaptation—decline—at 60-dB ANSI. Forty students with normal hearing (ages 18–45) were divided into four groups, tested at 1, 4, 6, or 8 kHz, with ICP vs SA as the within-groups variable. Results showed significantly greater adaptation in the ICP as opposed to the SA condition at 1, 4, and 6 kHz. There was no significant difference at 8 kHz. Thus mean differences between the two methods were greatest at the frequencies most closely coinciding with the meaningful frequency range of speech, as previously hypothesized [E. M. Weiler et al., J. Acoust. Soc. Am. 101 (1997)]. It is suggested that differences in subject behavior for speech and nonspeech frequencies may be modulated by differences in active processes of perception. These results will be discussed in the light of recent related work.