Break Expectancy and Uncertainty

Expectation of a break in timing seems strongly influenced by uncertainty about the break occurrence. The two experiments reported here suggest this to be the case whether uncertainty is varied by manipulating information on a global or a local level. Thus, in experiment 1, information was manipulated on a global level, because subjective probability about break occurrence developed over trials. In experiment 2, local information, a cue, was provided within each trial to inform participants of the break arrival. In both cases, longer intervals were produced when participants were more certain about the break, suggesting that pulse accumulation was affected.

Results from Fortin and Massé's (2000) study also support the hypothesized relationship between break expectancy and uncertainty. In that study, the strongest break location effect was obtained when all trials included a break, that is, when participants were certain that a break would occur (experiment 2; see results in Figure 9.2 of this chapter). The effect tended to weaken when trials with no breaks were included, hence reducing certainty about the break (experiment 4; see results in Figure 9.5 of this chapter). Furthermore, the effect was almost abolished when participants knew before starting the interval production that there would be no break in the current trial (cued condition of experiment 4; shown in Figure 9.5 of this chapter). In that condition where participants were informed that no break would occur, productions shortened drastically, suggesting that accumulation was relatively undisturbed. The opposite was obtained in the cued condition of experiment 2 reported here, where the cue had the exact opposite function, that is, made participants certain of the break occurrence. This presumably induced a complete or almost complete interruption in pulse accumulation (see Figure 9.10), hence longer productions in the cue condition (see Figure 9.9).

0 0

Post a comment