Break location ms

FIGURE 9.7 Results from experiment 1. Mean produced intervals, not including breaks (± SE), as a function of break location in trials with and without breaks, in conditions of high and low frequency of trials with breaks.

mean produced interval at each combination of break location, duration, and type of break signal. A mean produced interval was also computed in trials with no breaks.

Mean intervals not including break duration are shown as a function of break location in the high- and low-frequency conditions in Figure 9.7. In trials with breaks, mean intervals increased linearly with increasing break location in the high-frequency condition, R2 = .99, with a mean slope of 0.21 msec. The relation was strongly linear in the low-frequency condition as well, R2 = .94, but with a much lower slope, 0.08 msec. Slopes of productions as a function of break location in trials with breaks were computed for each participant. A t-test for independent samples showed that the mean slope was significantly greater in the high-frequency condition (M = 0.21, SD = 0.13) than in the low-frequency condition (M = 0.08, SD = 0.11), t(18) = 2.46, P < .05. These results show generally that the break location effect, attributed to expectation of a break, is stronger when trials with breaks are more frequent. This observation supports our main hypothesis: increasing the degree of certainty about break occurrence increases the level of expectancy for a break, hence strengthening the break location effect.

An analysis of variance (ANOVA) was performed on mean produced intervals not including breaks, in trials with breaks only, with one nonrepeated (break frequency) and two repeated (break location and type of break signal) factors. The interaction between break location and frequency was significant: F(3, 54) = 3.46, P < .05. The lengthening of produced intervals as a function of break location was stronger in the high- than in the low-frequency condition. The effect of break location was also significant: F(3, 54) = 15.91, P < .0001. Productions lengthened with increasing prebreak duration. Mean intervals did not differ significantly in low- and high-frequency conditions: F(1, 18) = 2.82, P > .05. No other effects were significant.

As shown in Figure 9.7, in the low-frequency condition, productions in no-break trials seem to follow the general linear trend present in trials with breaks. This is not the case in the high-frequency condition, where mean productions were even shorter in trials with no breaks than at the longest prebreak duration, 2200 msec, although a t-test for dependent samples showed that the difference was not significant (M = 3498.1, SD = 329.74 vs. M = 3393.9, SD = 349.97; t(9) = 2.00).

The results of experiment 1 confirm the break location effect found in previous experiments: participants produced longer intervals when they were expecting a break for a longer duration. This supports Fortin and Massé's (2000) conclusion that expecting the interruption of a signal being timed perturbs pulse accumulation, as illustrated in Figures 9.3 and 9.4.

The main finding here, however, is the interaction between break location and frequency of trials with breaks. The lengthening of produced intervals as a function of break location was more pronounced when the degree of certainty about the break occurrence was high. These results suggest a slower rate of accumulation during the prebreak period when participants are relatively certain that a break will occur. The slower rate of accumulation could result from an increase in frequency of attentional shifts. An illustration of the attentional-shift interpretation, showing pulse accumulation in both frequency conditions, is shown in Figure 9.8.

Accumulation starts on the first key press and proceeds with brief interruptions due to attentional shifts until the break occurs. Figure 9.8 shows the process for break location of 2200 msec. In the low-frequency condition, attentional shifts are assumed to be less frequent so that when the break occurs, the pulse count would be higher than in the high-frequency condition. After the break, accumulation resumes at its normal rate in both conditions until the criterion is reached, which

0 0

Post a comment