Casual Fridays: Those moving dots

Last week we asked our readers to identify the agents and actions in some point light displays. These displays show the motion of the joints of the actor (say, a human walker) and we are remarkably good at identifying various actors, actions, and even emotions.

Our three movies showed some kind of quadripedal action, and we provided a long list of possible anwers. Two of our three movies depicted very typical actions for the actor--a dog walking, and a human baby crawling. The third movie showed something slightly unusual--a human adult crawling. Did the unusualness of the action matter?


Here are the top responses we got as people viewed the movie showing an adult human crawling. For the most part, people successfully matched the actor and action, though a sizable group understood the action to be that of a gorilla. There is a significant interaction between certain kinds of experience and identifying this movie: parents were more likely to see this figure as human than dog owners. Dog owners are more likely to say "gorilla," and parents are more likely to mistake the crawling adult for a crawling baby. So, how do dog owners do with the actual dog movie?


Well, most folks do recognize the dog as a dog, with the two pig options a distant second. However, the parents are actually more accurate than the dog owners. Almost all the parents also had pets (not just dogs), which actually makes the difference more interesting: Something about being a parent seems to be adding to the ability to interpret action. The final movie was a crawling baby, something parents certainly dealt with at some point (though we didn't ask how old the children were).


So much for parenting and visual expertise! By far the most popular response to our crawling baby was "gorilla walking." In fact, both parents and dog owners are equally likely to identify the baby as a crab or a kangaroo! Stripped down to its essence, the action of a crawling baby is really tough to recognize.


More like this

I wonder if the crawling baby dots were mis-interpreted because of a trick of angle or perspective? The (rear) legs of the baby are mostly obscured, and the arms look much much longer from the point of view of the camera. Perhaps a true side view would get better accuracy?

And a nit-pick, you say after the first graph that "Dog owners are more likely to say "gorilla,"", but there's no way that's a significant effect... The parents clearly respond differently from the other groups, but not the dog-owners...

Thanks! Enjoy these surveys!


The angle might be part of it, though the movie was made from an actual baby. We thought about showing the hidden parts (back right leg), but that looked even weirder.

With the crawling adult, the dog owners are more likely to say "gorilla" than the parents (yellow bar significantly higher than green), but you are right that dog owners are not more likely than everyone (blue bar).

Is there anywhere that I can see all three animations again? I'd like to see them again knowing for sure what they are.

Even knowing that the animation above was of the baby, it took me a while to finally see it. My confusion was on the leg. The two dots representing the knee and the ankle, I saw as being the ankle and toes of something like a dog, and the hip as being the knee, and I thought that the hip was just absent.

Question - Is the speed of the baby's locomotion the actual speed, or has it been accelerated? From my own experience as a parent, the motion seems much faster than a baby usually moves. I suspect the speed of motion is the most deceptive factor for all participants.

By Chez Jake (not verified) on 03 Mar 2006 #permalink

No, it's not accelerated. All the animations are actual speed. Of course, depending on a baby's age, he or she can move at widely differing rates.

About the baby's speed--yes, that's how fast she was going (no acceleration in the movie). We were really surprised, because the actual video seems totally reasonable, and then the dots look amazingly fast. We didn't think to ask for permission to post the actual video from the parents. I think we just don't notice how fast the arms and legs are moving, because the whole baby isn't moving forward very fast.