To investigate how hearing loss of primarily cochlear origin affects the loudness of brief tones, loudness matches between 5-and 200-ms tones were obtained as a function of level for 15 listeners with cochlear impairments and for seven age-matched controls. Three frequencies, usually 0.5, 1, and 4 kHz, were tested in each listener using a two-interval, two-alternative forced-choice ͑2I, 2AFC͒ paradigm with a roving-level, up-down adaptive procedure. Results for the normal listeners generally were consistent with published data ͓e.g., Florentine et al., J. Acoust Soc. Am. 99, 1633-1644 ͑1996͔͒. The amount of temporal integration-defined as the level difference between equally loud short and long tones-varied nonmonotonically with level and was largest at moderate levels. No consistent effect of frequency was apparent. The impaired listeners varied widely, but most showed a clear effect of level on the amount of temporal integration. Overall, their results appear consistent with expectations based on knowledge of the general properties of their loudness-growth functions and the equal-loudness-ratio hypothesis, which states that the loudness ratio between equal-SPL long and brief tones is the same at all SPLs. The impaired listeners' amounts of temporal integration at high SPLs often were larger than normal, although it was reduced near threshold. When evaluated at equal SLs, the amount of temporal integration well above threshold usually was in the low end of the normal range. Two listeners with abrupt high-frequency hearing losses ͑slopes Ͼ50 dB/octave͒ showed larger-than-normal maximal amounts of temporal integration ͑40 to 50 dB͒. This finding is consistent with the shallow loudness functions predicted by our excitation-pattern model for impaired listeners ͓Florentine et al., in Modeling Sensorineural Hearing Loss, edited by W. Jesteadt ͑Erlbaum, Mahwah, NJ, 1997͒, pp. 187-198͔. Loudness functions derived from impaired listeners' temporal-integration functions indicate that restoration of loudness in listeners with cochlear hearing loss usually will require the same gain whether the sound is short or long.

This paper tests the hypothesis that the loudness ratio between equal-SPL tones with different durations is the same at all SPLs. Detection thresholds and levels required to produce equal loudness for 5- and 200-ms tones presented in quiet or in broadband noise were measured using adaptive, two-interval, two-alternative forced-choice procedures. Tone levels ranged from 5 dB SL to 90 dB SPL for the long tones and about 100 dB SPL for the short tones. Results from six listeners with normal hearing show that the amount of temporal integration, defined as the level difference between equally loud 5- and 200-ms tones, varies nonmonotonically with level and is greatest at moderate levels. The average amount of temporal integration in quiet is about 15 dB near threshold, increases to a peak of 27 dB when the 5-ms tone is about 58 dB SPL, and decreases to about 15 dB near 100 dB SPL. For masker levels of 40, 60, and 80 dB SPL, the amount of temporal integration near masked threshold remains near 15 dB. The maximum amount of temporal integration decreases as masker level increases and occurs at progressively higher levels. At high levels, the amount of temporal integration is nearly the same as in the quiet for all masker levels. Loudness functions derived by applying the equal-loudness-ratio hypothesis to the data yield excellent predictions of loudness matches between tones in the quiet and partially masked tones with the 40- and 60-dB maskers. For the 80-dB masker, clear deviations are present. These results support the equal-loudness-ratio hypothesis, but suggest that intense masking may alter the loudness ratio between 200- and 5-ms tones.