Huh? Preschoolers using statistics to learn psychology? Weren’t we taught that young children are egocentric–they can’t understand the world outside of themselves? Well a new study sheds light on how children learn preference, and it’s neither consistent with egocentrism nor affective cues (how others respond to preferred items), but instead statistics. That’s right–the human brain evaluates statistical information to come to conclusions. Go figure.
The study set out to determine how preschool-aged children “understand the actions of other people in terms of underlying psychological causes, such as desires, preferences, beliefs, emotions, and knowledge states.” It did so by creating an experiment testing a child’s recognition of nonrandomness, and whether this recognition allowed the child to make conclusions about another person’s preferences.
To do so, researchers recruited 72 preschoolers (average age: 4 years, 1 month) and put them in a room individually with a squirrel puppet named Squirrel. They then watched Squirrel choose toys out of a box. The available toys were of two types–foam red circles and blue flowers–in different proportions, randomly selected. The proportions were either 18%, 50% or 100% target toy (let’s say red circles). Squirrel then proceeded to pull out five samples of the target toy–still following?–and played with those toys for a short period. The toys were then put back in the box and the boxes removed. Out came three dishes with five samples of three types of foam toys–red circles, blue flowers and yellow tubes (as a control)–and the children were told to give Squirrel the toy it preferred.
Aha! Here’s where it gets interesting: The children gave the squirrel the correct toy–the target–most often when it was the least present, that is, when at an 18% ratio. When present at 50%, children gave the target correctly only half the time, and at 100%, they correctly gave the target only 25% of the time. Hmmm…
Statistical analysis of the results showed that when the target toy was in the lowest proportions, the results were significantly greater than a chance occurrence, whereas in the 100% proportion the results were no greater than chance.
This study shows that the developing brain uses statistical information to determine preference. If Squirrel chooses red circles despite their being less of them in the box than blue flowers, then it must prefer red circles. Right. When the proportions were equal however, or when there were only red circles, then the little brains could not infer which toys squirrel actually preferred. Right again.
Now this may seem obvious to some, and judging by some comments I’ve already read about the study this seems to be the case for a few people. However, to make this more interesting, the researchers set up a second study where they used affective cues–eyes widening, oohs and aahs, and so forth–when they received the preferred toy (in this study they used adults instead of Squirrel to test if children use visual and audial cues to make preference determination). And again the results showed that children more correctly chose the preferred toy when it was in lower proportion, despite affective cues. Genius.
I love this study because it gives some preliminary info on the workings of the young mind and how we come to certain understandings. Unfortunately, the same cannot always be said of the adult mind. Reading about this study initially in a TIME online article, I was appalled at the conclusion presented by the author,
“Even preschoolers, in other words, can see that some people might need more help getting what they want when less is available to them.”
This is a gross misrepresentation of the results and the author of the piece clearly did not understand the study. If you happen to come across this nonsense in any other article, please note that this study showed that children used statistical information to infer preferences in other people only. It had nothing to do with altruism, although I’m sure some people would love to see it that way to further their own beliefs. But this study was on learning mechanisms, and at that it was brilliant. We’ll leave the morality for another study.