The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 60:P102-P105 (2005)
© 2005 The Gerontological Society of America
Age-Related Differences in Multiple-Object Tracking
Lana M. Trick,
Tahlia Perl and
Naina Sethi
Department of Psychology, University of Guelph, Ontario, Canada.
 |
Abstract
|
|---|
Multiple-object tracking is the ability to attend (keep track of) the positions of multiple target items as they move among other items. The performance of young and older adults (M = 19 and 73 years old, respectively) was compared in two versions of a tracking task in which participants were required to monitor the positions of 1–4 moving targets in a field of 10 moving items. All participants were capable of tracking more than 1 item at once, but on average older participants tracked around 3 items at once whereas young adults tracked 4. Results suggest that there are age differences in the ability to either track or report positions of moving items. All participants reported positions of 1–4 targets in a field of 10 items with 100% accuracy when items were static.
MULTIPLE-OBJECT tracking is the ability to attend the positions of multiple target items as they move among identical items. Pylyshyn and Storm (1988)
found that young adults could track four to five target items simultaneously. Properties of tracked items are reported with greater speed and accuracy than those of other items. However, the advantage does not accrue to items that fall between targets—contrary to the predictions of a zoom lens model in which the attentional focus can be expanded to hold all the enclosed items (Sears & Pylyshyn, 2000). Moreover, neural activity during multiple-object tracking is different than would be expected if participants made discrete movements of attention between items (fMRI, Culham et al., 1998). Multiple-object tracking is considered an important component of attention, and the ability to monitor multiple-item positions is necessary for many day-to-day tasks (e.g., turning across traffic at a busy intersection when driving). It is consequently important to study this ability in older adults.
In this investigation, we used challenging versions of the task (10-item displays, 10-s tracking intervals) to provide stringent tests of what older adults could accomplish. We used full report (participants were required to report the locations of all targets) to make it easier to assess the number of tracked locations, even though this procedure makes extensive demands on working memory. Working memory demands may well be unavoidable in multiple-object tracking tasks. There is evidence of memory scanning even with partial report. Response times increase with the number of targets to be tracked at once, even when participants only have to determine if one specific item was a target, as would be expected if participants were serially scanning through a list of target locations in memory (Pylyshyn & Storm, 1988). There is controversy, but some have linked limitations in multiple-object tracking to visual–spatial working memory (Fougnie & Marois, 2004).
Enns and Trick (in press) suggest that one can best understand selective attention by using a framework derived from the combination of two fundamental dimensions: selection with and without conscious awareness (controlled and automatic), and selection by innate and acquired internally directed cognitive mechanisms (exogenous and endogenous). According to this framework, age-related differences are more notable in the controlled-endogenous mode of selection because brain areas mediating this type of processing are among the last to develop and first to deteriorate with age or pathology. Multiple-object tracking requires controlled-endogenous processing because participants have to deliberately select some items while ignoring others. There is certainly evidence that older adults exhibit diminished performance on other indices that involve this type of selection: attentional visual searching, dynamic searching, filtering based on motion, and marking locations of already attended (particularly moving) items (Faubert, 2002; Jiang, Luo, & Parasuraman, 2002; Kramer, Martin-Emerson, Larish, & Andersen, 1996; Watson & Maylor, 2002). However, some view selective attention as a variety of working memory (e.g., Tuholski, Engle, & Baylis, 2001), and there are also age differences in visual–spatial working memory (Jenkins et al., 2000). Either perspective predicts age differences in multiple-object tracking performance.
|
METHODS
|
|---|
Participants
Researchers recruited young adults from Guelph's university participant pool and older adults from a seniors' recreational center (age, M = 19 and 73 years, respectively). To ensure that both groups had comparable levels of health and visual function, researchers screened participants so that only healthy individuals with better than 20/40 Snellen acuity (Experiment 1) and no uncorrected visual disabilities (Experiment 2) were included. That left 20 young and 18 older adults in Experiment 1 and 20 young and 20 older adults in Experiment 2.
Materials and Procedure
Participants in both studies began by filling out questionnaires on their health. Those in Experiment 1 also completed the Snellen Acuity Test and the Sine Wave Contrast Test (Ginsburg, 1993), whereas older participants in Experiment 2 did the Early Treatment of Diabetic Retinopathy Scale Acuity Test, Pelli–Robson Contrast Sensitivity Test (Pelli, Robson, & Wilkins, 1988), and the Standardized Mini-Mental State Examination (Molloy, Alemayehu & Roberts, 1991). Results are summarized in Table 1.
In both studies, participants sat 45 cm from a Macintosh G4 Powerbook. The tracking field (the area in which items moved) was a central black rectangle occupying 22.96° x 17.33° visual angle on the computer screen. The items were 10 circular shapes, each with a diameter of 1.45°. These shapes were blue surrounded by a white border. Each item's rate of movement varied randomly from frame to frame, ranging between 0° and 9.35°/s (frame = 16.5 ms). Item motion was constrained so that items repelled each other and the borders of the tracking field; items could touch but not occlude one another or escape the tracking field.
In Experiment 1, participants tracked the position of one to four circles designated as targets. Each trial had five stages.
Stage 1: Initialization
Participants hit the "OK" button to initiate the trial. Ten circles were presented in random locations on the tracking field for 1105.5 ms.
Stage 2: Target acquisition
Over a period of 181.5 ms, 1–4 of the circles repeatedly flashed to indicate they were targets (181.5 ms off, 181.5 ms on for the duration). There was then a 495-ms pause with all 10 circles static.
Stage 3: Tracking
All items began to move randomly and independently of one another. Movement continued for 10 s.
Stage 4: Report
Items stopped moving and participants used the computer mouse to select the items that were targets. In every trial, participants were required to select as many targets as they had been required to track.
Stage 5: Feedback
The actual target items flashed over a period of 744 ms (181.5 ms off, 181.5 ms on for the duration).
After instruction, participants were given 8 trials of graded practice: 2 at each target numerosity, starting with the easiest (1 target) condition. There were 40 randomly ordered experimental trials.
In Experiment 2 we used a variant of the task in which the items were meaningful, if whimsical, figures: spies (targets) that "disguised" themselves as civilians (1.45° blue happy faces); see Figure 1. Targets revealed themselves by switching repeatedly from spy to civilian form during target acquisition. We made the following changes to the procedure: Target acquisition duration was reduced to 1,650 ms (targets alternated, 165-ms spy, 165-ms happy face for the duration); in report, participants pointed to the targets, and a research assistant selected target positions by using the mouse; feedback duration was increased to 1,260 ms to facilitate learning (165-ms spy, 165-ms happy face for the duration). As well, assistants gave a test of immediate visual–spatial memory before the tracking task. Participants underwent the target-acquisition phase, but they were then required to point at the targets immediately after the 495-ms pause. We measured the percentage of correctly identified targets.
|
RESULTS
|
|---|
In each trial, we measured the percentage of target items that were correctly identified. We analyzed the average percentage of correctly identified targets as a function of age (young adults, older adults) and number of targets to be tracked (one, two, three, or four).
In both studies, to determine the maximum numbers that could be tracked, we compared performance with that which would be expected if participants were guessing some locations. We derived expected outcomes for each number of targets by calculating the probability of randomly guessing the position of one or two of the targets (Freund, 1981, p. 181) and by using these probabilities to calculate the accuracy expected if some items were truly tracked and some were guessed. For example, if a participant was given the task of tracking three items at once, and only tracked two and guessed one location from the remaining eight positions, then his or her expected accuracy would be (2 + 1/8)/ 3 = 70.83%. These expected accuracies are plotted as dashed lines in Figures 2 and 3. To determine the maximal number of items that could be tracked, we used one-sample t tests to compare the observed accuracy with that expected if the position of one of the tracked items was guessed.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 2. Percentage of correctly identified targets by young and older adults when tracking 1–4 targets in a standard multiple-object tracking task with 10 moving items. Standard error bars are included. Dashed lines indicate expected accuracies if participants were randomly guessing the positions of 1 or 2 of the targets they were required to track
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Figure 3. Percentage of correctly identified targets by young and older adults when tracking 1–4 targets in the "catch the spies" variant of the multiple-object tracking task with 10 moving items. Standard error bars are included. Dashed lines indicate expected accuracies if participants were randomly guessing the positions of 1 or 2 of the targets they were required to track
|
|
Experiment 1
Young adults performed with almost perfect accuracy when tracking up to four targets at once (M = 98.2%), though accuracy dropped with the number of targets, F (3, 57) = 7.9, p <.001, 
2 =.29. The older adults also performed well (M = 90%), but the number of targets had a stronger effect on their tracking accuracy. Thus, age and number of targets had an effect on tracking accuracy, that is, age, F(1, 36) = 27.7, p <.001, 2 =.44, and number, F(3, 108) = 49.4, p <.001, 2 =.58, but there was also an Age x Number interaction: F(3, 108) = 26.9, p <.001, 2 =.43 (see Figure 2).
Younger participants performed far better than would be expected if they were guessing one of the locations, which suggests that on average they successfully tracked four items at once. In contrast, the older participants perform about as well as would be expected if they tracked three items and guessed the position of the fourth when required to track four items at once: t(17) = –1.3, p >.05. However, they performed significantly better than would be expected if they could only track two items and guessed the third when required to track three items: t(17) = 6.5, p <.001. Therefore, it seems that, on average, older participants successfully tracked up to three items at once.
Experiment 2
In the pretest, participants were required to point at the locations of targets immediately after target acquisition. Even though the duration of target acquisition was reduced to 1,650 ms, all participants recalled 100% of the positions correctly for one to four targets. This suggests that any observed differences in tracking performance were not due to age differences in immediate visual–spatial working memory or age differences in the required amount of encoding time.
Tracking accuracy was high for both age groups (M = 96.1% and M = 87.9% for young and older groups, respectively), but there were main effects of age and number as well as an interaction: age, F(1, 38) = 17.3, p <.001, 2 =.31; number, F(3, 108) = 42.3, p <.001, 2 =.53; Age x Number, F(3, 114) = 14.09, p <.001, 2 =.27 (see Figure 3). As in Experiment 1, younger adults performed significantly better than would be expected if they tracked three items and guessed the position of the fourth when required to track four: t(19) = 7.8, p <.001. Older participants performed significantly worse than would be expected if they tracked three and guessed the position of the fourth when required to track four, t(19) = –2.3, p <.05, though they perform better than would be expected if they tracked two items and guessed the position of the third when required to track three: t(19) = 4.3, p <.001.
Overall, participants were 2.5% less accurate in Experiment 2 than Experiment 1: F(1, 74) = 4.0, p <.05, 2 =.05. There were no Age x Experiment or Age x Experiment x Number interactions (F < 1 for both). In both experiments, on average, older adults successfully tracked around three items whereas younger adults tracked four.
|
DISCUSSION
|
|---|
Though the absolute number of items tracked may vary on the basis of the total number of items in the display, rate of motion, and so on, these studies reveal age differences in multiple-object tracking performance for large numbers of targets. There are several explanations, but some can be eliminated at the outset. Because older participants performed with almost 100% accuracy when tracking 1 item, the results cannot be explained by age-related insensitivity to motion or age-related inability to maintain concentration during a 10-s tracking interval. Results cannot be accounted for by age-related reductions in sensory or attentional resolution (He, Cavanagh, & Intriligator, 1997), because the number of items in the display was constant at 10, regardless of the number to be tracked. It seems unlikely that differences in tracking performance occurred because older participants failed to encode all the target positions during target acquisition, because all older participants were capable of reporting the positions of up to 4 items with 100% accuracy even with a reduced (1,650 ms) target acquisition period. Therefore, the explanation must lie either in age differences in tracking the positions of multiple independent items or maintaining their representations in working memory until they can be reported.
|
Acknowledgments
|
|---|
This project was supported by grants from Co-operators Insurance, the Natural Sciences and Engineering Research Council of Canada (238641-01), and CanDRIVE. Some of the results from Experiment 1 were presented at the annual meeting of the Canadian Society for Brain, Behaviour, and Cognitive Science (June 2004, St. John's, Newfoundland, Canada). Part of the data from the young adults group in Experiment 2 is presented in another paper (Trick, Jaspers-Fayer, & Sethi, 2004). With respect to the testing of the older adults in Experiment 2, we also thank our CanDRIVE collaborators: Evanne Casson, Heather Hollinsworth, Maria Muir, Lyne Racette, and Blair Nonnecke.
|
Footnotes
|
|---|
Decision Editor: Thomas M. Hess, PhD
Received for publication March 22, 2004.
Accepted for publication September 8, 2004.
|
References
|
|---|
- Culham, J. C., Brandt, S. A., Cavanagh, P., Kanwisher, N. G., Dale, A. M., Tootell, R. B. (1998). Cortical fMRI activation produced by attentive tracking of moving targets. Journal of Neurophysiology, 80,2657-2670.[Abstract/Free Full Text]
- Enns, J. T., Trick, L. M., in press A framework for studying age-related change and individual variation in selective attention. In Ellen Bialystok (Ed.), Lifespan cognition: Mechanisms of change. Oxford University Press.
- Faubert, J. (2002). Visual perception and aging. Canadian Journal of Experimental Psychology, 56,164-176.
- Fougnie, D., Marois, R. (2004, May). Are the capacity limit of VSTM and attentional tracking one in the same? Poster presented at the annual meeting of the Vision Science Society, Sarasota, FL.
- Freund, J. E. (1981). Statistics: A first course. (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall.
- Ginsburg, A. (1993). Sine Wave Contrast Test (Contrast Vision Test). Chicago: Visumetrics, Stereo Optical Company.
- He, S., Cavanagh, P., Intriligator, J. (1997). Attentional resolution. Trends in Cognitive Sciences, 1,115-121.
- Jenkins, L., Myerson, J., Joerding, J. A., Hale, S. (2000). Converging evidence that visuo-spatial cognition is more age-sensitive than verbal cognition. Psychology and Aging, 15,157-175.[Medline]
- Jiang, Y., Luo, Y., Parasuraman, R. (2002). Priming of two-dimensional visual motion is reduced in older adults. Neuropsychology, 16,140-145.[Medline]
- Kramer, A. F., Martin-Emerson, R., Larish, J. F., Andersen, G. J. (1996). Aging and filtering by movement in visual search. Journal of Gerontology: Psychological Sciences, 51,P201-P216.[Abstract]
- Molloy, D. W., Alemayehu, E., Roberts, R. (1991). A Standardized Mini-Mental State Examination (SSMSE): Its reliability compared to the traditional Mini-Mental State Examination (MMSE). American Journal of Psychiatry, 148,102-105.[Abstract/Free Full Text]
- Pelli, D. G., Robson, J. F., Wilkins, A. J. (1988). The design of a new letter chart for measuring contrast sensitivity. Clinical Vision Sciences, 2,187-199.
- Pylyshyn, Z., Storm, R. (1988). Tracking multiple independent targets: Evidence for a parallel tracking mechanism. Spatial Vision, 3,179-197.[Medline]
- Sears, C., Pylyshyn, Z. (2000). Multiple-object tracking and attentional processing. Canadian Journal of Experimental Psychology, 54,1-14.
- Trick, L. M., Jaspers-Fayer, F., Sethi, N. (2004). Multiple-object tracking in children: The "Catch the Spies" task. Manuscript submitted for publication.
- Tuholski, S. W., Engle, R. W., Baylis, G. C. (2001). Individual differences in working memory capacity and enumeration. Memory and Cognition, 29,484-492.[Medline]
- Watson, D. G., Maylor, E. A. (2002). Aging and visual marking: Selective deficits for moving stimuli. Psychology and Aging, 17,321-339.[Medline]
TOP
Abstract
Methods
