New research methods reveal that babies and young children learn by rationally testing hypotheses, analyzing statistics and doing experiments much as scientists do
Very young children's learning and thinking is strikingly similar to much learning and thinking in science, according to Alison Gopnik, professor of psychology and affiliate professor of philosophy at the University of California, Berkeley. Gopnik's findings are described in the Sept 28 issue of the journal Science. She spoke about her work in a video briefing with NSF.
New research methods and mathematical models provide a more precise and formal way to characterize children's learning mechanisms than in the past. Gopnik and her colleagues found that young children, in their play and interactions with their surroundings, learn from statistics, experiments and from the actions of others in much the same way that scientists do.
"The way we determine how they're learning is that we give them, say, a pattern of data, a pattern of probabilities or statistics about the world and then we see what they do," said Gopnik.
For example, in a series of experiments Gopnik and her colleagues used machines with the ability to light up and play music and asked young children to make them go.
"We found that like scientists, they tested hypotheses about the machines and determined which one was more likely," said Gopnik.
But before we rush to put toddlers on an earlier academic track, Gopnik's research shows that encouraging play, presenting anomalies and asking for explanations prompts scientific thinking more effectively than direct instruction.
"Everyday playing is a kind of experimentation--it's a way of experimenting with the world, getting data the way that scientists do and then using that data to draw new conclusions," said Gopnik. "What we need to do to encourage these children to learn is not to put them in the equivalent of school, tell them things, or give them reading drills or flash cards or so forth. What we need to do is put them in a safe, rich environment where these natural capacities for exploration, for testing, for science, can get free rein."
Gopnik's research was supported by NSF through the Social, Behavioral and Economic Sciences directorate. In her paper Gopnik described the work of Laura Schulz of MIT, also supported by NSF through the Education and Human Resources directorate. Schulz's studies show that children's play involves a kind of intuitive experimentation where they examine things and events to discover cause and effect underlying them. A video showing some of her experiments is attached. It was part of a paper she published in Science last year, 16-Month-Olds Rationally Infer Causes of Failed Actions.
More details about Gopnik's work are available in a TED talk she gave last year.