Stimulus Range And Modality Effects

Although, as reviewed above, there are numerous reports in the literature of modality effects in the milliseconds range, recently Melgire et al. (submitted) directly compared the influence of signal modality on temporal bisection performance in the seconds and milliseconds ranges. Each participant completed two test sessions on different days. For half of the participants, the first session anchor durations were 2 and 8 sec, and the second session anchor durations were 200 and 800 msec. The order of duration assignment was reversed for the remaining participants. Both auditory and visual probe durations were presented sequentially in random order in the same test session. Interestingly, although there was a robust effect of signal modality when 2 and 8 sec were used as the anchor durations, thereby replicating the results of Penney et al. (2000) showing that sounds are judged longer than lights, there was no effect of signal modality when 200 and 800 msec were used as anchor durations. One possible explanation is that the durations used in the milliseconds version of the task were too brief to allow a modality effect mediated by a flickering mode switch to be revealed. It is possible that when a timing signal initially captures attention, the switch is maintained in the closed state for some time before "flickering" develops. If this is the case, then for short-duration signals, i.e., milliseconds range, there may not be sufficient opportunity for the development of differential levels of flicker during the stimulus presentation. Consequently, the auditory and visual signals might not seem any different in the temporal bisection task when durations in the short milliseconds range are used. For longer signals (e.g., 2 vs. 8 sec), mode switch flicker develops and the modality effect is revealed as a shift in subjective duration that is proportional to the length of the probe duration. The major shortcoming with this explanation is that there is extensive evidence of robust modality effects with milliseconds range stimuli from a number of other timing paradigms (e.g., Goldstone and Goldfarb, 1964; Wearden et al., 1998). Perhaps there is something unique about the temporal bisection procedure in the milliseconds range, but this seems unlikely given that the data obtained by Melgire et al. (submitted) were typical of such procedures in other respects; i.e., there was an orderly increase in the percent "long" signal classification that assumed a sigmoidal shape (cf. Allan and Gerhardt, 2001; Allan and Gibbon, 1991; Penney et al., 1998, 2000; Wearden, 1991; Wearden and Ferrara, 1995, 1996); the point of subjective equality was observed to be close to the geometric mean; and functions superimposed on a relative time scale (Allan and Gibbon, 1991). Alternatively, the primary mechanism underlying the modality effect in the seconds range may be distinct from that involved in the milliseconds range. For example, milliseconds range modality effects may be primarily due to differential latencies to initiate and stop timing for the various signal modalities. Because onset-offset effects are absolute amounts independent of the duration of the timing signal, they would make a larger contribution to the size of a modality effect for a milliseconds range signal than they would for a seconds range signal. Indeed, as Grondin (1993) noted for filled vs. empty duration comparisons, timing onset-offset differences may wash out with longer durations. Of course, even if this is true, it does not explain why a modality effect was not obtained in the 200- vs. 800-msec condition of Melgire et al. (submitted).

0 0

Post a comment