One recommendation to reduce COVID-19 transmission is to keep a distance of at least 2 meters/6 feet from others. If you are a pedestrian making your way through busy city streets, this advice is easier said than done. There’s a lot to keep track of to maintain distance with people coming from different directions and at different speeds. Some are walking dogs, some are carrying packages and shopping bags, some are walking on the left side and some are walking on the right side, some are walking in groups spanning the walkway and some are walking in groups in single file, some are wearing masks and some are not wearing masks. Getting from point A to point B can feel like playing a game of frogger in real life (remember Frogger? See the screenshot on the left below for a refresher).
This effort isn’t entirely dissimilar to that of a radar operator. The task faced by an operator of marine radar is one that requires to detect all sorts of obstacles, such as refugee boats, marine life, pirate ships, landmasses, etc. (see the screenshot of a radar screen on the right above). The operator’s perceptual limitations could lead to any number of costly consequences – from loss of human life to loss of expensive military hardware.
If perception is actually a rate-limited process, then how can this rate be measured? If one spreads attention to multiple objects, does divided attention also divide the rate at which one becomes aware of any given object? These questions were addressed by Lappin, Seiffert, and Bell (pictured below) in research recently published in the Psychonomic Society journal Attention, Perception, & Psychophysics.
Participants watched a dynamic “naval air warfare” display containing multiple moving objects (representing planes) with a stationary ship in the center. The set size of moving objects in the display varied from 1, 2, 4, to 6 across trials. Participants tried to rapidly detect changes in either color or motion direction that could occur with equal probability to any object in the set. About 15 changes per minute occurred unpredictably during ongoing 6-min observation periods. The main question was how the speed of detection would depend on the number of objects being monitored by the participants.
The participants engaged in a detection task and a selective task. In the simple detection task, well-trained participants responded quickly, by pressing a spacebar, to any of these display changes. And in the selective detection task, they responded to a subset of the changes that reflected a “threat” to the ship in the center.
The 30-second video shows the experience of a participant. When you watch the video, count how many display changes occur.
[videopress 6UIAuTfz]
How many changes did you detect? (answer: 8). If you watch it again, you can keep track of the changes by reading the counter, called “ChScore,” in the upper right corner. The video shows changes in color, motion direction, and size, which slightly differs from the video used in the current study, where the objects’ sizes remained constant.
To measure how the detection of visual changes develops over time, the authors computed hazard rate functions. Hazard rate functions measure the rate of the detection process (in bits/sec) at any given time if the display change has not already been detected at that time.
The figure below shows these measures of the time-course of detection in each of the 16 conditions (4 set sizes x color vs. motion-direction changes x simple vs. selective detections). As shown in the figure, the detection rates were influenced by all of the experimental conditions. In all conditions, however, the detection rates rose and fell in similar fashion within a time-window of about ¼ second. Thus, display changes were detectable at about the moment they occurred, but then disappeared from awareness.
A key finding was that when attention was spread over a larger set size of moving objects, then detection rates were divided by essentially the same amount across all conditions and also constant over time. Similar constant effects of divided attention were also found in a preceding study (on the capacity of awareness) with different stimulus information and different performance.
Several analyses showed that the detrimental effects of divided attention were approximately constant over time and over variations in stimulus and task conditions. One such analysis is indicated in the figure below, which quantifies the cumulative total information (in bits) accumulated over time at three successive times (600, 700, and 800 ms) after a display change in each of the experimental conditions. The figure also compares the actual and predicted cumulative information if the three experimental variables (set size, color vs. direction changes, simple vs. selective detections) had independent influence on the detection process rate at any given time. Importantly, the predicted effects of the attentional set size were the same across all of these conditions and times.
The authors interpret the results as suggesting
… that visual perception operates at a limited rate, as a channel with limited capacity.
In further communication, Lappin stated,
… the meaning of these results — involving a limiting rate of conscious perception, and the effect of this capacity limit on perceptual and cognitive information processing — is quite relevant to everyday experience and performance. The findings also have practical applications in interfacing humans with information technology.
Psychonomic Society article featured in this post:
Lappin, J. S., Seiffert, A. E., & Bell, H. H. (2020). A Limiting Channel Capacity of Visual Perception: Spreading Attention Divides the Rates of Perceptual Processes. Attention, Perception, & Psychophysics, 82, 2652–2672. https://doi.org/10.3758/s13414-020-01973-9
