音の高さの弁別に要する反応時間

抄録

An attempt was made to measure reaction time in the discrimination of two pure tones as a function of pitch difference and to examine Takada's hypothesis proposed from the view point of information theory (Jap. Psychol. Res. 1960). The subject was instructed to press the telegraph key on the right hand as soon as possible, if he judged that the second of the two successive tones was higher in pitch than the first, and to press the left key if he judged lower. Fig. 1 shows a block diagram of the apparatus used to present the stimuli and to measure reaction time. Reaction time was defined as the time between the on-set of the second tone and the key-pressing response. The frequency of the standard tone was constantly 1100cps. throughout all experiments, while that of the comparison tone was varied as described later. The loudness of tones was approximately 70phon. Four subjects were used.
In the preliminary experiment, the effects of the tonal duration and the interstimulus interval on reaction time were examined using five different durations, ranging from 250 to 2000msec, and six intervals from 60 to 2000msec. The frequencies of the comparison tones were 1070 and 1130cps. The mean reaction times obtained in this experiment are shown in Fig. 3 as a function of inter-stimulus intervals. Each line indicates the condition of duration.
In the main experiment, twelve comparison tones were used, frequency of which varied from 310 to 2370cps in 12 steps, excluding the same frequency as the standard. From the results of the preliminary experiment, 0.5sec duration of tones and 1.0sec inter-stimulus interval were adopted. The results of this experiment are shown in Fig. 5 in which each point plotted represents the mean of eighty reaction times (4 subjects×20 repetitions) obtained for each condition of the transmitted information which was calculated from the absolute pitch-differences in logarithmic mel, using the following equation:
I_t=log₂M-log₂Δm (1)
where I_t is the amount of information transmitted by stimulus patterns consisting of two tones, M is the maximum range of pitch in mel and Δm is the pitch difference in mel between the standard and the comparison tones. The lines in Fig. 5 indicate the predicted values from the equation,
Rt=aI_t+b (2)
where Rt is the reaction time in millisecond, and a and b are constants. The analysis of variance shown in Table 1 indicates that the effects of frequencies (F), session (S), and subjects (Ss), and interactions, Ss×S, Ss×HL (Higher versus Lower groups of tones as compared with standard), F×Ss×HL and Ss×HL×S were all statistically significant beyond the 1% level, while F×S was significant beyond the 5% level. The practice effect was shown only between the first and the later three sessions, in which the reaction time was shorter than in the beginning. The average rate of information processing was calculated to be 22.7bits/sec from the obtained data by the least square method. This value is of the same level as those of Takada and some others, but differs from that of Henmon.
In the supplementary experiment, the simple reaction time without discrimination task was measured under the same condition as the preliminary experiment. The effects of duration and interval in this experiment show the same trend as the preliminary experiment.
The results of this study supported Takada's hypothesis on the information processing in discrimination.