The effect of psychophysical method on the loudness of continuous and interrupted pure tones

Using the Bekesy tracking method for loudness judgement tasks such as most comfortable loudness (MCL) and recalled loudness (RL) measurements, normal listeners have tracked continuous (C) tones at lesser intensities than interrupted (I) tones. The resulting continuous/interrupted (C/I) separations have ranged up to 22 dB. Explanations of the unexpected C/I separations have not been in agreement. The purpose of this study was to investigate whether psychophysical method has a significant effect on the C/I separation. Subjects were six normal-hearing adults, each of whom participated in three practice and ten test sessions of one hour each. Thirty-six experimental conditions were presented. The test signals were C tones, 200/200 msec I tones and 200/800 msec I tones, all at 1 kHz. The reference intensities wer 20, 50 and 80 dB SPL. The four methods were Bekesy tracking method (BTM), method of adjustment (MAdj), method of limits (ML) and method of constant stimuli difference (MCSD). Mean values for Point of Subjective Equality and C/I separation were calculated. C/I separations obtained with the BTM were significantly larger than C/I separations obtained with any other psychophysical method. These results demonstrate that C/I separations obtained during suprathreshold Bekesy tracking tasks by normal hearers are largely caused by the measurement method itself rather than by differences in C and I tone loudness. Time order errors could explain the BTM effect.