Multiple Target Tracking: Does practice make perfect

1
Working Memory and the visual tracking of
multiple moving targets
Roy Allen, Peter McGeorge David Pearson,
Department of Psychology, University of Aberdeen,
Scotland. Tele 44 (0)1224 272387 Fax 44
(0)1224 - 273426 email roy.allen_at_abdn.ac.uk

• INTRODUCTION
• The working memory (WM) model (Baddeley Hitch,
  1974), has been successfully applied to many
  visuospatial phenomena. But, as yet, not to the
  perception and monitoring of moving objects
  within the external environment.
• A dual-task paradigm consisting of a
  computerised multiple-target tracking task (e.g.
  Pylyshyn Storm, 1988), together with various
  secondary tasks believed to tap specific WM
  resources, may give some insight into the process
  of tracking multiple moving targets.
• METHOD
• Subjects
• 144 post- and undergraduate students, 73F and 71M
  took part, aged between 16 and 47 (mean 21.37, SD
  4.7). All had normal or corrected-to-normal
  vision. Each was paid a small honorarium.
  Participants were allocated to one of four
  secondary task conditions visual, auditory,
  spatial or verbal. For half, these began before
  target acquisition the remainder after.
• Materials
• Six blocks, each of 40 trials, were prepared in
  advance. Each trial began with a centralised
  solid white fixation square on a black
  background. After a delay, ten static crosses ()
  appeared, distributed randomly (see fig. 1). A
  number of these, designated as targets (1-5),
  were flashed on and off several times the target
  acquisition phase (TA), the remainder were
  distractors.
•  

• Procedure
• Participants were tested individually in a
  darkened room. Each completed the single and dual
  task conditions. Every trial began by pressing
  the spacebar. It was emphasised that they should
  keep their eyes on the fixation square throughout
  each trial. Further, they had to pay attention to
  the indicated targets. If, and only if, a target
  was probed, they had to press the keyboards
  spacebar as quickly as possible. It took 2 hours
  to complete 200 trials. Accuracy and RTs were
  recorded.
• The secondary tasks, carried out during each dual
  task trial, were labelled
• Visual - categorizing single digits, flashed at
  1/sec in the fixation square, as high (6-9) or
  low (1-4)
• Auditory - categorizing tones, played at 1/sec,
  as high or low
• Spatial tapping the four corner buttons of a 3
  x 3 matrix once/sec
• Verbal repeating the once/sec.
• RESULTS
• The number of targets that participants could
  successfully track was obtained by subtracting
  the number of false alarms (FA)from the number of
  hits (as per the high-threshold model of signal
  detection, e.g. Snodgrass Corwin, 1988). This
  figure was then compared to the number of misses,
  using binomial tests.
• Results suggest that, in the single task,
  participants can always track 4 targets
  significantly better than chance, p lt 0.05 (see
  Fig. 2). During the dual task, for the Visual and
  Auditory conditions, participants can always
  track 3 targets significantly better than chance,
  p lt 0.01 (see Fig. 3), whilst for the Spatial and
  Verbal conditions, participants can still track 4
  targets significantly better than chance, p lt
  0.05 (see Fig. 4).
• Analysis of secondary task data confirmed that
  participants always allocated adequate and
  sufficient resources, always performing above 85
  accuracy.
• Other analyses on RTs, response bias and
  detection sensitivity were carried out but are
  not reported here.

Figure 3 Summary of all tracking task
performance in dual task condition, by secondary
task secondary task commencing before TA. Filled
markers indicate number of targets successfully
tracked in each case.
Figure 4 Summary of all tracking task
performance in dual task condition, by secondary
task secondary task commencing after TA. Filled
markers indicate number of targets successfully
tracked in each case.
DISCUSSION Whatever the secondary task,
performance never significantly varied with
secondary task onset. This suggests that the TA
phase of the experiment is not associated with WM
processes and supports Pylyshyns claim that
target acquisition is preattentive. TT, on the
other hand, does seem to utilise WM. Tracking
performance in the Visual condition suggested a
role for central executive resources, an idea
proposed by Yantis (1992), but decrements might
also be due to primary and secondary tasks
sharing the same modality. Tracking performance
in the Auditory condition, though, refuted this
latter interpretation. However, the tracking
performance decrement might just be the effect of
performing any two tasks simultaneously.
Tracking performance in the Verbal condition did
not support this notion. Finally, a tracking task
might seem to be very spatial in nature, however
tracking performance in the Spatial condition did
not support this either. An interpretation is,
therefore, that the central executive plays a
role in the tracking of multiple moving objects.
However, a similar pattern might be obtained
independently of central executive input if, in
fact, visual indexes were amodal. REFERENCES Badd
eley, A.D. Hitch, G.J. (1974). Working Memory.
In G.A. Bower (Ed.), Recent advances in learning
and motivation, Vol. VIII. (47-90) New York
Academic Press. Pylyshyn, Z.W. Storm, R.W.
(1988). Tracking multiple independent targets
Evidence for a parallel tracking mechanism.
Spatial Vision, 3 (3), 179-197. Snodgrass, J.G.
Corwin, J. (1988). Pragmatics of measuring
recognition memory - applications to dementia and
amnesia. Journal of Experimental
Psychology-General 117(1) 34-50. Yantis, S.
(1992). Multi-element visual tracking Attention
and perceptual organization. Cognitive
Psychology, 24 (3), 295-340.

Figure 2 Summary of all tracking task
performance in the single task condition. Filled
marker indicates number of targets successfully
tracked