Perception of Duration with Irregularly Filled Intervals

The perception of interval duration can be distorted by a multitude of factors related to stimulus characteristics. One of them is the filled-duration illusion, where filled intervals are perceived to be longer than their empty counterparts. The filled-duration illusion has been shown to occur with intervals consisting of one continuous signal (e.g., Wearden et al., 2007). Most notably, however, the illusion also occurs with intervals containing several short signals (“fillers”) in between a beginning and an end marker (e.g., Thomas & Brown, 1974). Only few investigations have looked at the effect of the filler signal characteristics on the perception of duration. The present work focused on the role of regular versus irregular spacing of filler signals in interval duration judgments. Stimuli were auditory filled intervals between 0.5 and 1.5 s duration, marked by five 10ms 1000Hz filler tones. Five different levels of irregularity in the timing of the fillers were defined, ranging from regular (that is, spacing between every two neighboring fillers is equal) to 50% of the regular spacing. In a two-interval forced choice task, participants judged which of two intervals appeared longer. One interval was always regular, while the other one was irregular. The serial position of the physically longer interval as well as the irregular interval were counterbalanced and randomized. Psychometric functions were fitted to the proportion of responses plotted as a function of the difference in duration. The point of subjective equality and the just noticeable difference were estimated. While the variation in irregularity did not affect duration discrimination sensitivity, results indicate that irregular intervals are judged to last shorter than regular ones of the same physical duration and this effect increases with the amount of irregularity. Furthermore in trials where the regular interval precedes the irregular one, irregular intervals were specifically judged shorter than their regular counterparts, when the first interval between fillers was significantly longer than the second one. These findings suggest that duration perception of filled intervals can be affected by the properties and regularity of the individual filler stimuli, opening the road to a new interpretation of the reasons behind the filled-duration illusion. Implications for a computational model of interval duration perception are discussed considering the perception of filler stimulus timing to be affected by the expectation built up from previously presented stimuli.