Dr. Rachel Severson New Psychology Study

Kids, Confidence and Credibility

UM psychologists add to research on how children learn

By Courtney Brockman

Eliminate those filler words and convey certainty when speaking to children because true confidence is key for relaying information to kids, says a new study co-headed by a University of Montana developmental psychologist.

Dr. Rachel Severson, an associate professor of psychology, published the findings this past January in collaboration with Susan Birch at the University of British Columbia and Adam Baimel at Oxford Brookes University. Based on a series of three studies with 662 children at UBC, the experiment discovered kids’ preferences on learning from confident versus hesitant individuals.

Severson, whose interest in working with children grew while witnessing their problem-solving in Nepal and India, researches how kids think about others’ minds. She earned her doctorate in developmental psychology at the University of Washington and arrived at UM in 2016 after time at the University of Oslo, Western Washington University and UBC.

She says the new study suggests children are more savvy at learning from others than previously thought. Children don’t just prefer to learn from confident people; they also avoid learning from overly confident people — making them less likely to believe misinformation and ensuring they are learning accurate information.

The unusually large sample size of kids gave the researchers a high power to detect differences for the study. Some of the children were part of the Early Development Research Group’s decades-old database of 30,000 families. Others entered the study through public engagement at the UBC Living Lab in the Telus World of Science, or Science World.

Severson says parents and children alike took to the experiments.

“One of the wonderful things about doing research with children is they love having an attentive adult who’s just listening,” Severson says. “A lot is set up as games, so it’s like we’re playing a game together. Kids love it, and parents love seeing what the kids do and hearing about what we are researching.”

Previous research shows children even as young as 2 1/2 years prefer to learn new information from confident individuals over hesitant ones. But is that tied to whom they think is more knowledgeable? What about someone who is confident but not actually knowledgeable?

That is what Severson and her colleagues wanted to measure.

In the first study, children ages 3 to 12 watched videos of two models — one confident and the other hesitant — who provided different answers to the same question. The children were asked to select which answer they thought was correct and whom they thought was smarter, and their pattern of choices provided insight into which model they thought was more credible.

To convey confidence versus hesitancy, the informants used linguistic cues like “I know” versus “I think.” And they conveyed how certain they were with paralinguistic expressions such as rising intonation or a physical expression like shrugging shoulders.

In the condition where both models were knowledgeable about the information, but one person was confident and the other hesitant, the children preferred the confident individual — as expected.

In the other “uniformed condition,” neither model was knowledgeable, although one responded confidently and the other hesitantly. Children age 3 to 5 preferred to learn from the confident model, even though her confidence was not justified.

“What we found was kids, especially younger kids, were still swayed by the confidence,” Severson says, “but they did not account for how well the model’s confidence was calibrated with her knowledge.”

But children age 5 to 12 years showed a significant move away from the unjustifiably confident model and toward the hesitant model whose level of confidence matched her knowledge.

Two follow-up studies assessed how kids react to two confident people, one informed and one not. As expected, kids up to age 8 preferred a confident and knowledgeable model, showing they attend to both knowledge and confidence.

But with two paired hesitant individuals, one knowledgeable and one not, there was no preference in children up to 8 years.

Why? Severson thinks children learn whom to trust in a phased approach. Up to about age 5, they trust confident individuals. Then around age 5 they start showing skepticism of unjustifiably confident people. But learning to trust people who are justifiably hesitant is harder, and not until around age 8 do kids begin to recognize that hesitancy is justified when a person lacks knowledge.

How children understand the internal states of others interests Jingjing Sun, a cognitive psychologist and assistant professor of educational psychology at UM. The study, she says, suggests a stepwise delay in 4-year-olds, who are capable of calibration for confidence but not yet for justifiable hesitancy.

“But that they were able to find children as young as 4 in the study able to distinguish people who are overconfident from people who are truly confident is pretty fascinating to me,” she says.

Sun studies older children, from about third to sixth grade, focusing on their capacities for critical thinking, relationship skills, cooperating, decision-making and understanding others’ perspectives in a school setting.

In a study on emergent leadership done in Chinese classrooms, Sun asked students to rate their peers’ talkativeness, as well as whom they perceived as friends, in questionnaires before the study. She found children who are talkative and of a higher social status were more likely to emerge as leaders among their peers in small groups given open-ended problems. However, their leadership was not necessarily effective.

Younger children focused on external, observable qualities like physical appearance when asked to describe their confident friends. But at age 5, children started focusing more on internal qualities and dispositions.

“Kids, especially middle-age kids, have the capacity to understand that other kids’ thinking is different from their own,” Sun says. “However, they may not use such capacity to solve real-world issues.”

Sun applies her research through collaborating with teachers in Missoula elementary classrooms on how school learning can be better structured to support this developmental stage of children’s ability to think from others’ perspectives.

Holding moral dilemma discussions — such as whether a grizzly bear who attacked a hiker in Yellowstone should be put down — encourages kids to share their positions and ideas and challenge one another’s thinking in rich, thought-provoking discussions, she says.

“The result is very rewarding,” Sun says. “One of the teachers came at the end of the study and told me she never had that opportunity when she was an elementary school kid to participate in much discussion.”

In conducting more studies with children, Sun says it’s important to consider how having siblings may affect children’s cognitive skills in judging confidence and knowledgeability in others. She says accounting for cultural appropriateness while looking at research data is also vital.

Interestingly enough, the tendency to trust overconfident people is still evident in adults.

A preliminary study Severson conducted with a diverse group of undergraduates at UBC, including international students, showed unjustifiably confident people are still trusted by adults.

“What we’re seeing is that about roughly 30% of adults are still swayed by that confident individual, even when they are not knowledgeable,” Severson says. “That’s helpful for us to know what’s the ceiling level.”

One follow-up study at UM measured how contexts with moral implications such as resource allocations, harm and injustices change children’s responses to confident and hesitant people.

Children around age 5 start preferring hesitant informants if the moral stakes are high, such as scenarios involving hungry animals and a shortage of food.

“What may be happening is kids are interpreting confidence as a kind of rush to judgment in the morally relevant scenarios, whereas hesitancy may indicate appropriate deliberation,” Severson says.

Severson, who has studied children’s interactions with robots and other “personified” technologies, also will bring smart speakers into the mix to find out whether kids will treat them as trustworthy sources for learning factual information, versus information with moral implications.

What does all this mean for parents?

Since children around age 5 begin taking the internal state of a person into consideration, as well as their confidence, Severson says they can recognize when their parents don’t know information.

“It’s good for parents to pay attention that their kids will view them as being credible, moreso and increasingly as they get older, when they’re appropriately confident or hesitant,” Severson says. “It’s really important for parents to acknowledge when they don’t know something their kids ask about.”

In a challenging and uncertain situation like a pandemic, parents should prioritize good communication and connection with their kids, as well as sometimes just admitting when they don’t have the answer.

“I think it’s important to have that transparency and to be well-matched to what you are putting out,” Severson says.

Missoula Library Adds UM Living Lab

Soon parents can observe their children’s learning in action at the UM Living Lab in the new Missoula Public Library.

UM has joined the City of Missoula and Missoula Public Library to build robust STEM education and career pathways. This $1.3 million initiative is the first of its kind in the nation to be funded by a Science Education Partnership Award supported by the National Institutes of Health.

With community and Flathead Reservation tribal partners, the collective effort is creating a two-story exhibit called the DNA Tornado — a helix for young learners to scale — a suite of health science programs and the UM Living Lab.

The UM Living Lab, part of the National Living Laboratory initiative, is the “expansion of an idea,” said Severson, who will direct the lab. It will include developmental science research and outreach, as well as engage the community in public health, neuroscience, genetics and more — all highlighting research done at UM and in the wider community.

Free and open to the public, the nationally funded initiative is projected to reach over 500,000 people per year, and it is tentatively scheduled to open in late 2020. •