Generality Across Stimulus Conditions And Time Estimation methods

The effect of expecting a break appeared in the two experiments of the present report in spite of differences in stimulus conditions relative to Fortin and Masse's (2000) experiments, where the effect was first observed. Whereas the break period was indicated by an interruption in tone presentation in previous experiments (see Figure 9.1), new stimulus conditions were used here, with the break period marked by a tone presentation (see Figure 9.6) in the continuous break-signal condition of experiment 1, and by two tones bounding the break period in the discontinuous signal condition. The break location effect did not vary across stimulus conditions, which was also observed when different stimulus conditions (visual vs. auditory, filled vs. empty breaks) were manipulated within experimental sessions, in a within-subject design (Fortin and Bedard, 1999).

We have also observed effects of break expectation in time discrimination studies (Tremblay et al., 2001; Tremblay and Fortin, submitted). In one experiment, two tones of 2.5 and 3.0 sec were used as short and long durations, respectively. Participants were first familiarized with the two target durations. In each of the following experimental trials, one of the two tones was presented with a break in tone presentation. Break location was varied: it could occur when 30, 50, or 70% of the tone duration had elapsed. Break duration, as well as break location, was varied from trial to trial. The stimuli used in that experiment are illustrated in Figure 9.11. The task was to classify the total tone duration (prebreak plus postbreak) as corresponding to the short or long target duration by providing a "short" or "long" response.

A shortening of perceived duration with increasing prebreak duration was suggested in responding to both short and long durations: the proportion of correct responses increased with increasing prebreak duration in "short" trials, whereas it decreased in "long" trials. Overall, participants responded short more often as pre-break duration increased, a result compatible with previous data obtained in discrimination experiments when location of a nontemporal task was manipulated (Buffardi, 1971; Casini and Macar, 1997).

The break location effect was large and reliable: substantial differences in proportion of short responses were observed with varying break location. More importantly, when the effect was tested over a wider range of location values, the function relating the proportion of short responses to prebreak duration was not linear, but sigmoid. Thus, differences in proportions of short responses were reduced at extreme values of break location, that is, when two values of break location placed near the

0 0

Post a comment