Why is it that higher income kids tend to score better on achievement tests than poor kids, even at the youngest ages? One explanation from the 1990s is the so-called 30-million word gap, in which researchers observed how higher income parents talked to their kids more and estimated that low income kids heard 30 million fewer words before the age of four.
But a new generation of researchers has been questioning whether the quantity of words really matters. Two studies in 2014 found that the quality of interaction between parents and babies was a better predictor of language skills. A 2018 neuroscience study calculated that back-and-forth conversation was more important to brain development than the sheer number of words.
Now scholars are dissecting early childhood speech even more and analyzing the content of these back-and-forth conversations. Thanks to advances in wearable audio recorders and natural language processing technology, it’s become more practical to “listen” to hours of speech inside homes.
One pair of researchers, Kathleen Corriveau of Boston University and her former graduate student Katelyn Kurkul, now at Merrimack College, have produced a series of studies that delve into how some parents are answering their children’s questions and how those answers might make a difference in how we learn. The studies are small and haven’t been replicated but their work is an interesting glimpse into this new direction of early childhood research.
“Prior research has looked at frequency, how often parents and caregivers give explanations,” said Kurkul. “But there hasn’t been a lot of emphasis on how these explanations affect children’s learning.”
Corriveau and Kurkul’s most recent study of more than 100 four and five-year olds at Boston area preschools and a museum found that children who were exposed to more “mechanistic language” learned more than children who received shorter, less detailed explanations.
For example, if a child asks the question, “Why is there a battery?,” a simple, non-mechanistic answer could be, “To make it work.” That’s not wrong but it’s not very informative. A more detailed “mechanistic” response supplied by one parent in the study was: “The battery is there to give the toy power. When the batteries are connected to these buttons, then the toy will work because it has power.”
Measuring learning among young children is tricky. In this case, the researchers had children play with a snap circuit toy of many pieces that could be assembled to turn on a fan or a light bulb.
Children were just as likely to reassemble a fan system that was presented to them at the start of the exercise regardless of what type of explanation they received. But children who had heard the mechanistic type of explanation were twice as likely to assemble a new light bulb circuit that they hadn’t previously seen. That required them to understand the basics of a circuit and that the toy pieces had to be touching to turn the light on. The benefits of mechanistic explanations were true for both low- and middle-income children in the study.
“The mechanistic language is causing them to retain something about the concept that allows them to generalize it to a novel task,” said Kurkul.
In a previous study, the researchers noticed that children prefer non-circular answers to their questions. For example, a circular response to the question, “Why is the sky blue?” would be “Because it’s not another color.” When asked to choose between two adults, one who utters circular answers and one who doesn’t, kids tend to choose the person who offers the better non-circular reasoning, the researchers found. It didn’t matter if the answer was scientifically wrong; children seem to appreciate the attempt at a deeper explanation.
Kurkul and Corriveau have compared the answers that low-income and middle-income caregivers typically offer children and they found stark differences. Low-income parents and caregivers were more likely to give circular answers to children’s questions than middle-class parents and caregivers. And now the researchers believe, from the latest snap circuit study, that the better answers are not just preferred but also leading to better learning.
With more cities offering early childhood education, Kurkul’s research is now shifting toward improving teacher quality in these classrooms. She’s studying aspiring teachers to see what kinds of explanations they typically give and looking into how future teachers can be explicitly trained to avoid circular answers that don’t enlighten.
“It’s a challenge,” said Kurkul, “because you are bombarded with questions during these preschool years: why, why why? But those questions are leading to learning. So the explanations you provide really do matter.”
This story about mechanistic explanations was written by Jill Barshay and produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for the Hechinger newsletter.