Your car tells you what to see: Driving affects distance judgments
Wednesday, November 4, 2015
One of the most important functions of perception is to help organisms navigate through their environments. Different animals navigate through very different environments: think of birds flying at thousands of feet above the ground, bats catching moths in pitch darkness mid-flight, whales crossing entire oceans, bees finding nectar-rich flowers, monkeys scampering through dense tree foliage. The variety of perceptual systems that make this possible have evolved to be tightly integrated with the animal’s capacity for action.
But while most animals are limited to perceive in act by their evolved bodies, humans have created a dizzying array of tools that augment and extend our capacities for both perception and action. Microscopes and binoculars help us see things that are too small and too distant, telephones allow us to talk across spaces our voices cannot span, levers and pulleys allow us to lift weights that would otherwise snap our muscles, and cars allow us to travel at speeds far in excess of those our perceptual and motor systems evolved to handle—speeds that our not-too-distant ancestors thought would literally suffocate us. Yet, it is remarkable to consider how capably our perceptual and motor systems appear to cope with traveling at these speeds.
In an article recently published in Psychonomic Bulletin & Review, researchers Moeller, Zoppke, and Frings asked whether driving experience might augment how people evaluate distances in front of them because cars allow us to travel differently (much easier and faster) through space— than we would otherwise.
Participants—19-52 year old drivers— were randomly assigned to one of three groups. In the driver group, participants sat inside a stationary car and evaluated distances to traffic cones placed 4-20 meters in front of them. To evaluate the distance, participants instructed an experimenter to move two other perpendicularly oriented cones until they spanned the same distance as that between the participant and the target cone (see image c below). Participants assigned to the first pedestrian group performed the same task while seated in a chair (a). Participants assigned to the second pedestrian group performed the same task while looking through an aperture designed to simulate a car windshield to control for the possibility that any differences in evaluating distances between the driver and pedestrian group could be ascribed to peering through a frame.
After performing this set of judgments, participants assigned to the two pedestrian groups went a 10 minute walk. Participants assigned to the driver group went on a 60 km/h 10-minute drive. After this period, participants in all groups performed the distance evaluation task one more time.
Evaluating distances in this way tends to result in substantial underestimates of true distances. For example, participants have the cones positioned about 14 m from each other when the true distance to the target cone is 20 m. Strikingly, participants assigned to the driver condition underestimated the true distances by a much wider margin, for example estimating that the cone 20 m away was only 11.25 m away. This relative underestimation was also present when participants simply made their judgments through the car window. The underestimation grew larger after the drivers returned from their 10-minute drive (they now judged the 20 m cone to be 10.9 m away). In contrast, the pedestrians’ judgments were unchanged by having gone for a walk.
What might cause this effect? The authors reasoned that “entering a car does more to you than making transportation more comfortable. In fact, the moment you sit in your car, your (distance) perception of the environment seems to adapt to your new ‘action potential’ ”. Because a car allows us to cross the same amount of space much more easily, we end up perceiving the distance as actually shorter.
Theories invoking ‘action potentials’ of this sort have been strongly criticized over the past several years, however, (e.g., see here, and here for a response) and it is worth considering whether the present results can be explained in a different way. An alternative commonly suggested by critics is that participants in such tasks conform to their intuitive biases (or the demands of the experimenter). Perhaps participants think that they should under-estimate distances when they sit inside a car.
To test this possibility, I conducted a survey on Amazon Mechanical Turk, asking people to imagine looking through a windshield: would things in front seem closer than they actually were? Farther? Of 30 people, exactly 2/3 thought that objects would appear exactly the “right” distance away, 20% thought objects would appear farther than they actually were. Only 4 people (13%) thought objects would appear closer. When asked why, some participants ventured that perhaps the glass had a lensing effect. Such results certainly do not rule out an explanation based on experimenter demands, but they also do not lend it support.
What other mechanisms might explain Mueller and colleagues’ finding? Another possibility is that the observed effect is a consequence of the car effectively becoming an extension of our body, causing the 2 or so meters of the dashboard and hood in front of us to be discounted when making distance judgments. This would help explain the fairly constant increase in underestimates experienced by the driver group, but it would not explain why the underestimation was further exaggerated by actually driving.
Another possibility is that driving experience genuinely changes how we perceive distances because we have associated cars with forward movement such that our visual system applies a movement “prior” even when the car is stationary. An example of this kind of process at work is our perception of digital displays like those in the image below.
Most digital displays we see have a slight rightward tilt. Our visual systems appear to renormalize the tilt, making a right-tilting display to look more vertical. As a result, the digits in the image on the right looks substantially more tilted to people than the image on the left (both are tilted at exactly 12 degrees from the vertical). Similarly, if looking through a car window is generally accompanied by forward motion, distances may become (slightly) foreshortened.
If looking through a car windshield makes things appear substantially closer than they would otherwise, shouldn’t we notice? In the same survey that asked participants about looking through the windshield, they were also asked which car mirror(s) distort distances such that objects are actually closer than they appear. The correct answer—the passenger-side mirror—was chosen by 80% of participants. But 73% thought the driver-side mirror also distorted distances, and 33% thought the rear-view mirror did as well. If we take these results seriously, we would conclude that participants fail to perceive in the course of daily driving that the same objects look considerably farther in one mirror than another. This may seem like a gross failure of perception until we realize that if one of the goals of perception is to help us navigate, then an object a certain distance away should be perceived that distance away—inasmuch as that’s possible—even in spite of what it “looks like” in the mirror. Similarly, the reason we may not notice distances changing when we look out of a car window is that that is what the world looks like through a car window!
Article discussed in this post:
Moeller, B.; Zoppke, H. & Frings, C. (2015). What a car does to your perception: Distance evaluations differ from within and outside of a car. Psychonomic Bulletin & Review, DOI: 10.3758/s13423-015-0954-9.
Dr. Dionysus Lardner, Professor of Natural Philosophy and Astronomy at University College, London, reportedly remarked in 1830 that "Rail travel at high speed is not possible because passengers, unable to breathe, would die of asphyxia”