Researchers have known for some time that people are surprisingly accurate at visually judging distances to objects as far as 25 meters away. If you're allowed to briefly look at an object up to that distance away, then blindfolded, you'll walk right up to it with great precision. If you walk halfway, you can throw a ball the remaining distance, again, quite accurately.
But in 2000 Marla Bigel and Colin Ellard attempted a simple replication of the study: instead of viewing the object, volunteers were led blindfolded to the object and back, and asked to walk back to the object again. Now, instead of accurately walking the distance, they systematically overestimated the distance to the object. Could our feet deceive us more than our eyes, even when we're simply asked to retrace a path we've just taken?
There's another possible explanation: maybe being led is what causes the deception. In a new study, Ellard and Sarah Shaughnessy asked 30 volunteers to walk blindfolded along a roped-off 10-meter pathway. They could use the ropes to guide themselves, but were never led by researchers. As before, they walked the distance to the object, then returned to the starting point, and finally attempted to walk the same distance again. Another group of volunteers simply looked at the object and then tried to walk to it blindfolded. Here are the results.
Now everyone was very accurate at pacing off the distance, whether they had walked it off or viewed the target first. It appears that the act of leading a person blindfolded over a distance systematically skews their perception of the distance, but when they walk unassisted, people are able to accurately judge distances, even while blindfolded.
So if both viewing the object and walking to an object result in accurate distance estimates, then which method takes precedence? In a new experiment, Ellard and Shaughnessy had volunteers both walk to the object and view it -- but the object was a different distance from the viewer each time. So the object might have been 6 meters away when the distance was walked, but then it would be viewed 8 meters away. As you might expect, responses were biased towards the most recent distance, whether the object was viewed or walked to. But responses were biased more to the viewed object. When the viewed object was seen last and 8 meters away, viewers were just as accurate as when they had only seen the object at that distance.
So humans appear to be more comfortable using the visual system to judge distance than using the motion of their bodies. Researchers had initially found that that people were very bad at judging distances when blindfolded, but then realized that people were preoccupied with worries about whether they'd bump into a wall -- so all the experiments discussed here were conducted outside, in an open field.
Ellard, C.G., Shaughnessy, S.C. (2003). A comparison of visual and nonvisual sensory inputs to walked distance in a blind-walking task. Perception, 32(5), 567-578. DOI: 10.1068/p5041
Question: when walking to an object to estimate a distance, wouldn't most people count their steps? Might it be that being led messes up one's normal gait, so that, say, 15 steps while led is only 37 feet, while 15 steps while unled is 45 feet?
PS. I always find these reports of yours quite fascinating!
Good question: The researchers had the participants repeat rhymes to themselves while walking so they couldn't count their steps.
Perhaps related to the first question: It mentions the path is roped off. Could the subjects be "counting off" along the roped path?
Still seems like you could use the rhymes as a counting device though. Or did they have a mechanism for foiling that, too?
In order to repeat a task blind folded you have to rely on your memory of either how much time the task takes or how many steps. If you say a rhyme you can still use the subconscious memory of your journey to feel when you are back to the starting point. In order to combat this you could have the subjects walk at different speeds and gaits for each pass.
Some people estimate and walk these distances very frequently (for example professional tennisplayers). Will they be more accurate in estimating distances than others. Is this something that can be learned and improved by practice?
One possible flaw in this is that the subjects with the ropes were most likely holding them in their hands, which provides another sensation upon which to base the distance.
The researchers had the participants repeat rhymes to themselves while walking so they couldn't count their steps.
I suspect this would not work for some people. In one of Feynman's books he mentions how he found out that people have two different mental counting systems. This was in the context of timing something. Many people counted by using their talking subsystem, but Feynman did the counting by using his visual system (mentally picturing the numbers flipping by) and thus was able to keep counting while talking or listening.
I'm probably breaking some kind of unwritten rule here by commenting on this work since I did the studies, but I can't resist chiming in, first of all to thank Dave for the well-written and accurate summary of some of our work and secondly to address the interesting discussion about step-counting. The commentators are basically right in saying that there is no completely bulletproof strategy for preventing step-counting because if you design a really elaborate and difficult distractor you can make it hard for a person to walk -- if the task is too hard participants will grind to a complete stop! On the other hand, we and others have found that trying to step count in the face of a mild distractor (it was a while ago but I think Marla used a recitation task and Sarah used a backwards counting by an awkward interval -- like seven -- task) makes your performance quite a bit worse. Those who struggle hardest to arrive at some clever heuristic for "cracking" our task always seem to do worse than those who simply and automatically try to walk towards a mental image of the target. This is not to say that some kind of internalized counting, timing or measurement is not going on -- if it wasn't then we wouldn't be able to do this task -- and it's that measurement process we'd like to understand a bit better.
Thanks for commenting on this post -- you're not breaking any rules that I'm aware of.
It's fascinating to hear your anecdotal observation that step-counting doesn't really seem to work anyways. That makes a lot of sense to me -- just as we don't think about every muscle movement when we bring a glass of water to our mouths, we don't count steps when we try to go from one place to another -- we just think about going there and we're on our way.
In my opinion, a rythm could also be used as a mechanism for measure by relying on beats. It'd be much better have them tell you a story, or do mathematical operations while getting to the object.
I think the most interesting part of this study might be considering it's application to other forms of instruction. One could consider the result on the second part of the study as showing that people learned better when they led themselves than they did when the were led by another.
I don't believe that this really has anything to do with feet. When you go to a location by car and someone else is driving you tend to have a less accurate idea of how to get there (assertion anecdotal from personal experience) than if you were the driver.
As far as the subjects counting their steps i would think that asking them not to combined with mild distraction would suffice.
i trip all the time so maybe that's my problem