Influence of methodological factors on estimates of spectral loudness summation at medium and high levels

Verhey, J.L. and Appell, Jens-E. and Hohmann, V.
Equal-loudness-levels for 1000 ms bandpass-noise signals spectrally, centered at 2 kHz are measured in normal hearing listeners as a function, of the bandwidth in the range from 200 to 6400 Hz using an adaptive, 2-interval forced-choice 1-up 1-down procedure. The influence of the, following methodological factors is investigated: (i) the procedure, using interleaved tracks in comparison to a simple adaptive procedure, (non-interleaved), (ii) the level difference between test and reference, signal at the beginning of an adaptive track (starting level) and (iii), the bandwidth and level of the reference signal. When non-interleaved, tracks are used, a significant influence of the starting level is shown,, which is not observed in the interleaved procedure. Although an, influence of the reference bandwidth occurs in some subjects, no effect, is shown on the average data. The average spectral loudness summation, derived from the data is in good agreement with data from the, literature. However, a significant interindividual variance is observed., The results of the present study indicate that an adaptive procedure, with non-interleaved tracks is not appropriate to quantify loudness, summation. Instead, interleaved tracks should be used to avoid, influences of methodological factors (at least for group data). The, observed interindividual variance of the loudness summation data needs, further investigation (e.g. comparison with data from loudness scaling, or magnitude estimation).
01 / 1999