Estimation of short time intervals under different experimental conditions

A complex experiment is carried out in order to study the human ability to estimate and reproduce short time intervals in the 100-1000 ms range presented with light stimuli. The estimation task consists in verbal nomination of the time intervals presented to the subject in random order after preliminary training. In the production task the subject is expected to press a telegraph key simultaneously with the end of the light interval presented the duration of which is announced to him in advance. In the reproduction task the subject is to press the key a given time after the onset of the stimulus which duration is of 2 sec. The results show that in spite of the relatively good achievements in tasks involving verbal estimation of the duration of the intervals, the reproduction tasks are performed with systematic errors. In the production task the mean values of the response times decrease upon increasing of the required interval, while in the reproduction task they increase parallel with the increase of the required interval. Consequently, one and the same short time interval is perceived differently depending on the task to be performed, namely: verbal nomination, coincidence between hand movement and the end of the interval, or reproduction after verbal nomination.