Physical Time

Physical Time

FIGURE 10.2 A schematic for predicting how changes in attentional demands between training and test will affect young and older adults' performance on interval timing tasks. When attentional demands are greater at training than at test (top), pulse accumulation will occur more rapidly during the test trials than during training for the same unit of physical time. As a result, the number of pulses accumulated will reach criterion for response too soon. The opposite pattern occurs when attentional demands are greater during test trials than during training. In both cases, effects will usually be exaggerated for older adults because of their reduced attentional control.

adults' clocks too fast or too slow? Older adults' self-reports indicate that they often feel as if time passes more quickly than it did when they were younger (Fraisse, 1984; Schroots and Birren, 1990). This suggests that aging may lead to a slowing of the internal clock, which would make it seem as though external time were passing by quickly. Such slowing would be consistent with general findings that older adults are slower than young adults on many cognitive tasks (e.g., Salthouse, 1996). The idea that older adults have a slower internal clock than do young adults would at first seem consistent with ideas about age differences in controlled attention. That is, older adults' reduced attentional functioning might lead them to accumulate pulses more slowly than do young adults, resulting in a lower — and slower — clock reading.

Surprisingly, a meta-analyis of interval timing experiments using young and old adults found the opposite pattern: older adults had higher ratios of subjective time to objective time than did young adults, a result that at first might seem to suggest that older adults have a faster internal clock (Block et al., 1998).

However, as Block et al. (1998) note, this counterintuitive finding may largely result from the testing procedures used in the studies included in their review. Almost all of these experiments used absolute judgments of empty intervals. That is, participants either assigned a verbal label (e.g., 3 sec) to an experimenter-produced duration or attempted to produce the appropriate duration in response to the experimenter's query (e.g., "Hold this button down until you think 3 seconds has passed"). Within the context of the interval timing models described above, these procedures can be described as asking participants to compare current accumulator values acquired in the experimental setting to reference memory values acquired in daily life.

When asked to time an empty interval in the quiet setting of a laboratory, there is little to pay attention to except for time. Therefore, the attentional switch should remain steadily closed, allowing the accumulation of pulses that measure time. In contrast, daily experience is usually filled with many attentional demands and distractions in addition to time. To the degree that these demands and distractions disrupt attention to time, the function of the attentional switch will be disrupted, and fewer pulses will accumulate than would be the case in the lab. The discrepancy in attentional demands — and therefore pulse accumulation rates — causes the clock to "run faster" in the experimental setting than it does in everyday life.

Absolute time judgment tasks lead to overestimation and underproduction because they ask participants to compare clock readings acquired during the experiment, with a "fast clock," to values associated with labels learned in everyday life, with a "slow clock." Because older adults are more vulnerable to distractions and demands on attention than are young adults, they have a larger discrepancy between experimental-setting clock speed and everyday-life clock speed. This larger discrepancy for older adults leads to their showing exaggerated subjective/objective duration ratios compared to young adults, as shown in the studies reviewed by Block et al. (1998).*

The direction of age differences in subjective/objective duration ratios for absolute time judgments can easily be reversed by simply increasing the attentional burden within the experiment relative to everyday life. One way of doing this is to divide participants' attention within the experiment between timing and some other task. To the degree that the divided attention task presents attentional demands that are greater than those of the external environment, subjective/objective duration ratios should be less than 1.0. Because older adults are more vulnerable to demands that reduce the amount of attention available for timing, they should show a greater shrinking of subjective/objective duration ratios than do young adults under these conditions.

Craik and Hay (1999) tested young and older adults' interval timing performance under divided attention conditions using both estimation and production timing tasks. Rather than making time judgments about empty intervals, as was the case in the studies reviewed by Block et al. (1998), participants in these experiments made judgments about the characteristics (shape, size, brightness, etc.) of visual stimuli during the to-be-timed interval. In the estimation task, participants performed the visual judgment task until interrupted by the experimenter and asked to estimate how much time had passed. In the production task, participants were asked to perform the visual judgment task for a particular duration and to stop when they thought that duration had passed. In both cases, participants had to divide their attention and efforts between timing and the visual judgment task.

The increased attentional demands of the procedures used by Craik and Hay (1999) led to results that were very different from those reviewed by Block et al. (1998). In the Craik and Hay experiments, both age groups underestimated and overproduced time during the experimental trials, leading to subjective/objective duration ratios of less than 1.0. The distortion was relatively mild for young adults (with an overall duration ratio of 0.81), but dramatic for older adults, who had an overall duration ratio of 0.54. That is, older adults reported that about 33 sec had passed when the actual duration was 60 sec. This pattern is the exact opposite of the one found in the studies reviewed by Block et al., where older adults overestimated and underproduced time more than did young adults.

The contrast between the studies reviewed by Block et al. (1998) and those conducted by Craik and Hay (1999) fits nicely with the proposed attention-based framework for understanding older adults' interval timing performance, and demonstrates how divided attention manipulations can be used to test this framework. Both sets of experiments involved absolute judgments of time, where reference memory values obtained during the divided attention conditions of everyday life were compared with current accumulator values obtained in the experiment. The difference between them lies in the attentional demands present during the experiment. When there was little available in the experimental context to distract attention from timing, older adults tended to overestimate and underproduce time in the experiment (Block et al., 1998). In contrast, when the experiment required them to divide attention between the timing task and another task, older adults showed the opposite results, now underestimating and overproducing durations to a greater degree than did young adults (Craik and Hay, 1999).

* Young adults had subjective/objective duration ratios slightly over 1.0 in many of the experiments reviewed by Block et al. (1998), indicating that they, too, were affected by the difference in attentional demands between the outside environment and the lab. However, the deviations from 1.0 were typically much smaller than for older adults, and often not statistically significant.

Despite this general agreement, one aspect of Craik and Hay's (1999) findings might be seen as contradicting the ideas that dividing attention can affect interval timing performance, and that divisions of attention are especially detrimental for older adults. Contrary to their initial predictions and to previous findings (e.g., Brown, 1985; Fortin and Rosseau, 1998; Fortin, this volume), Craik and Hay did not find that timing accuracy decreased as the complexity (and thus attentional demands) of the visual judgment task increased, except at their longest duration (120 sec). They suggest that the lack of task complexity effects is attributable to a problem in their experimental design: all trials of the judgment task were 10 sec long, and participants may have used the consistent duration of the individual trials to help them keep their time judgments consistent across increasing levels of complexity. This explanation might be adequate to explain the general lack of task complexity effects, but is less acceptable as an explanation for the lack of age differences. Even in the 120-sec condition, where there were significant main effects of task complexity, older adults' timing performance was not more affected by task complexity than was that of young adults.

The resolution to this apparent contradiction may lie in the results for the visual judgment task. Although timing performance did not decline as visual judgment complexity (i.e., the number of decisions about shape, color, etc., that had to be made) increased, performance on the visual judgment task itself did. These complexity-related declines in performance were especially pronounced for older adults. This pattern of results parallels that of Li, Lindenberger, Freund, and Baltes (2001), who found that older adults preserved walking performance rather than memory performance when the difficulty of the two tasks was varied. In the Craik and Hay (1999) experiments, the preservation of timing performance in the face of increased visual judgment complexity may have occurred because participants prioritized the timing task and devoted their attentional resources there, rather than to the visual judgment task. This explanation receives some support from the results of a relative time judgment experiment (Vanneste and Pouthas, 1999) in which participants divided their attention between multiple to-be-timed stimuli, rather than between the timing task and multiple nontemporal stimuli, as was the case for the Craik and Hay experiments. Here, timing errors increased with the number of stimuli to be timed, and the increase in errors was more pronounced for older adults.

Are older adults' clocks too fast or too slow? The answer from these absolute judgment tasks seems to be "it depends." More specifically, the answer depends on whether the demands for controlled attention are greater in the outside environment or in the experimental setting. When test trials are conducted under full attention conditions, attentional demands will typically be less than in the outside environment, leading to quicker pulse accumulation and thus a fast clock. When test trials are conducted under divided attention conditions, attentional demands may be greater than in the outside environment, leading to a flickering of the mode switch, slower pulse accumulation, and thus a slow clock. Both of these effects will be exaggerated for older adults because they are more vulnerable than are young adults to demands on controlled attention.

The results for absolute timing judgments therefore fit nicely into the framework described earlier and pictured in Figure 10.2. Importantly, they do not show that older adults simply have slower clocks, as might be predicted from a speed-of-processing perspective; they also do not show a nonspecific increase in errors, as might be predicted from general memory or performance problems. Instead, the direction of older adults' timing errors depends specifically on the attentional demands of the experiment vs. the outside environment.

Furthermore, older adults' interval timing errors meet an important criterion for establishing them as the result of problems with attention or other aspects of clock speed: they are amenable to feedback. As described earlier, sensitivity to feedback is considered a hallmark of clock speed effects, because participants can use feedback to adjust to the new clock speed and relearn the number of pulses associated with the target duration (Meck, 1983). Wearden et al. (1997, experiment 4) asked both young-old (age 60 to 69 years) and older-old (age 70 to 79 years) adults to produce 1-sec intervals. Without feedback, participants overproduced the duration, and this overproduction was more pronounced for the older group. However, after receiving feedback (a display after each trial indicating the actual duration they had produced), participants became highly accurate and maintained this accuracy even after feedback was discontinued. This pattern strongly suggests that the older adults in this experiment were able to use feedback to adjust to the different attentional demands — and thus faster pulse accumulation and clock speed — of the experiment vs. daily life.

Absolute time judgment tasks are easy for participants to understand and have a great deal of face validity. However, they are problematic for investigating attention's role in interval timing because the experimenter does not control the attentional environment in which the target duration is learned and established in reference memory. It is therefore hard to pinpoint a priori whether the controlled attention demands are greater inside the experimental context or the everyday environment, and to what degree. Therefore, attention-based explanations of older adults' absolute time judgment performance, including the framework described in this chapter, are necessarily post hoc. Because of this problem, many investigations of older adults' interval timing use relative time judgment tasks, where the experimenter controls the demands for controlled attention at both training and test.

0 0

Post a comment