“How do flowers bloom?” “How do we have a heart?”
To adults, these questions sound simple, maybe even silly. To children, they unlock the world, representing the very beginnings of one of the most crucial developments in childhood: the growth of language. For Kathryn Leech, Ph.D., an assistant professor at the UNC School of Education, that development and growth form the building blocks of child literacy.
In her research, Leech is committed to uncovering the power behind these parent-child interactions. Her work focuses specifically on language and literacy development in Pre-K children and how they develop these foundations socially, primarily in a family or home context.
Activities such as mealtime, playtime, and bedtime may seem like normal family routines, but according to Leech, these contexts are pivotal moments for language acquisition. She has continued to investigate and work with families to improve these interactions throughout her academic journey — from graduate studies at the University of Maryland to postdoctoral research at Boston University and Harvard University to now at UNC-Chapel Hill.
As the recent recipient of the National Science Foundation’s (NSF) Faculty Early Career Development (CAREER) Program Award, Leech will further her research to optimize parent-child interactions in scientific learning through book reading, a common context that exposes and initiates conversation between parents and children.
This award will also enable her to interact directly with a population traditionally underrepresented in research – children and their families from rural communities – thus broadening our understanding of literacy and promoting family involvement in science education.
Learning through language
Leech’s research journey began as an undergraduate at Kenyon College, where she initially considered a career path in public policy or education policy. It wasn’t until she worked with an Early Head Start preschool in rural Ohio to fulfill service-learning requirements that she realized her interest in language development during early childhood. Observing how children learned language in the classroom inspired her to study what is known as “language-mediated learning” or the use of language to facilitate learning, memory, and problem-solving in social settings.
Building on these early experiences and interests, her work has helped advance an interdisciplinary approach to the field, emphasizing that it is not just decoding text and comprehension that impacts literacy. It’s the everyday interactions that caregivers have with children that help build a foundation for future reading success.
“That’s where I really got interested in these questions about what promotes language development,” she said. “How can we work with parents to encourage them to promote their child’s language development?”
Leech’s research methods typically involve her and her team members directly observing families in lab and home environments, gathering a language sample of a child and parent during an activity like interacting with a toy or reading a book. This transcript is coded, and language is categorized based on certain influential factors for the child’s language learning. From there, the researchers design interventions, such as encouraging questions and providing additional conversation prompts for caregivers, to measure the effectiveness of those interactions in language development.
One of the main factors Leech measures in these conversations is the presence of decontextualized language, a form of language that is not grounded in the present. These conversations can look like the parent referencing something that happened in the past or something that will happen in the future, as well as more complex topics, including explanations about how things work.
For example, Leech created and has studied the impact of using R.E.A.D.Y. — or Recall past events with your child; Explain unfamiliar concepts and words; Ask questions; Discuss future events; You can make a difference in your child’s future academic success — to encourage parents to incorporate certain types of language into their daily interactions with their children, such as recalling past events, explaining new words and concepts, asking lots of questions and discussing the future. This acronym is designed to be parent-friendly and easily rememberable to encourage the parent to use more decontextualized, complex language to increase their child’s literacy.
“What we found in our work is that parents’ questions directed to children support child language development because they encourage the child to talk more,” she explained. “So when the parent asks something like ‘Why were you so sad?’, that opens up the conversation, and the child answers with more than one word and encourages back-and-forth conversation, an experience we know is very helpful for language learning during this age.”
Capitalizing on curiosity
Leech classifies her research on the role of parent-child interactions within three interrelated themes. First, she investigates the contexts and specific aspects of these conversations that foster language and literacy development. Second, she aims to develop interventions to enhance literacy and learning so parents can have more complex, authentic conversations with their children.
Her goal is not to try to change what families do but rather to reinforce the positive practices already present in their interactions.
“We develop interventions and call them parent-focused interventions. The idea is based on research indicating that providing parents with information about these types of conversations, as well as building their self-efficacy and their confidence as an important teacher for their child, is a way that we can encourage parents to engage in more of these types of conversations,” she said. “We test whether we can provide parents with this information and whether that then corresponds to an increase in their child’s language and literacy development.”
Her third theme focuses on how children learn in these interactions, more recently in the area of scientific thinking. This research interest began while transcribing parent-child conversations in graduate school and realizing how children learn more than just vocabulary and complex language, but also the foundations of abstract scientific concepts, a type of contextualized language. This final theme has encompassed the majority of what Leech is focusing on now in her research at UNC, measuring topics from how children learn about electricity to how blood is pumped in the body.
“Three- to five-year-olds ask a ton of questions. They’re extremely curious, and this is what we consider to be a prime period to capitalize on that curiosity, to encourage parents to ask questions and provide explanations. We think this is a way to get children interested in science.”
Focusing on authenticity
Much of Leech’s research in early childhood scientific learning comes from observing conversations and designing interventions that are authentic and relevant to each family dynamic and culture. Central to this approach is moving away from formal learning contexts, like museums and classrooms, to focus on interactions at home.
During her post-doctoral studies at Boston University, Leech studied children’s conversations with parents around exhibits at the Boston Museum of Science. This mutually beneficial relationship allowed her to gather research and for the museum to gain insights into the exhibits’ effectiveness.
However, Leech realized that studying these conversations in this space was not gathering an accurate representation of all families living in the area. Museums are often situated in cities and may be financially inaccessible to some families.
“I wanted my current work to focus specifically on the home because we know that, for kids this age, science is everywhere in their everyday lives,” Leech said. “How can we understand how these conversations are occurring in the home between caregivers and their children, particularly for rural families where there may not be a museum that’s easily accessible?”
Being in the home also allows Leech and other researchers to expand their methodology by gathering a larger, therefore more representative, sample of a parent-child interaction. With technologies like the LENA Recorder — a small, child-safe recording device worn by children and known as the “talk pedometer” — they can record research subjects throughout the day, providing a more comprehensive snapshot of a child’s life and interactions.
“Because children’s language and literacy development is affected by many factors, having multiple measures often gives you a richer picture of what’s going on,” she said.
In Leech’s future research with the NSF Grant, she also plans to collaborate with the parents themselves in crafting specific interventions through co-design workshops, making sure the conversations are authentic to the families participating in the research.
“We realize that we want the families that we adapt the interventions for to understand: ‘Is this relevant to you? What do you value?’ and then make our adaptations accordingly.”
Following is a Q&A with Leech, who talked about plans with the NSF Career Award, collaboration with MIT in developing an AI tool, and where her work is headed. Answers are edited for space and clarity.
What does this NSF CAREER Award mean to you? What will you be able to do with this award, and how will it impact your research and our understanding of parent-child literacy?
I was excited because this was an opportunity for me to align all three of my research themes together in one project, which hadn’t happened before.
Another aspect I am excited about is that in addition to a research plan, all NSF CAREER projects have to include a plan to influence education in some way. My particular plan is focused on the Human Development and Family Science (HDFS) undergraduate major in the School of Education to develop and help to formalize more undergraduate research participation.
The students with whom I’ll be working will be working on my research project. At the same time, I hope to contribute to a more formalized track that students can take, getting more students interested in research earlier in their time here at Carolina. When we get them interested in their sophomore year, they have time to build up their experience and set them up for potentially going on to graduate school or whatever their next step is.
I went to a college where there weren’t graduate students, so the research experiences were focused on undergraduates. I’ve always had undergraduates in my research group, and they’re super important to the work I do, so I think this is very rewarding to be able to contribute hopefully to the HDFS program to make something that is sustainable.
You also have a collaboration with colleagues at MIT that involves developing an AI-powered app to aid parents during book reading. Could you share more about how this app works and its potential impact? How do you see AI affecting your future research on child literacy?
I think the most important thing to say is that I see AI as never replacing the adults in these interactions. I think we’ll learn so much more in the coming years about how children learn from these AI-powered technologies. The availability of this technology is so new and widely accessible, and this principle really guided the design of this project. We wanted to see how an AI-powered conversational agent can support or augment conversations between the child and their parent.
In this project, we designed an AI-powered conversational agent that is used during a book-reading interaction. The parent reads a book to the child, and the conversational agent is designed to listen for marker words in the text. When it hears a word, it chimes in with a question to encourage the parent and child to have a conversation using R.E.A.D.Y. talk. We have tested this out in a couple of different studies and found that it does work and that for the most part, parents and children enjoy the experience. It increases the talk during book reading, and the children are significantly more engaged when they’re reading with this agent.
We just finished a study where we measured the child’s vocabulary learning and found relative to kids who didn’t read with the agent, reading with the agent was associated with greater gains in vocabulary. So we’re seeing some learning from these experiences.
We also are excited because we built the original architecture of this agent in 2019, without any of the large language models that are now available. Our team and MIT, which is responsible for the software development part of it, are making really good headway into a second version that incorporates large language models like OpenAI so that we can do a lot more than what we were able to do initially.
Where do you see your research headed in the next few years?
I’m really interested in scaling my work more than I have early on. I’m still relatively early in my career, and so I think now that I have accumulated some of the evidence for some of my findings, I’m excited to develop interventions that have much more of a wide-scale impact than some of the proof-of-concept work that I’ve done previously.
I think technology comes into play as a potential method to disseminate this type of information in a much wider way, for example through phones and texts. I can see providing interventions through conversational agents as a scaffold where we don’t require having a human to train parents or to coach parents like we currently do.
Whether that is effective, that’s an empirical question, but I can see us moving in that way. The reason why is not to do it because it’s cool, but to do it because we can make a bigger impact on more people given the resources that we have.
References
Leech, K., Wei, R., Harring, J. R., & Rowe, M. L. (2018). A brief parent-focused intervention to improve preschoolers’ conversational skills and school readiness. Developmental Psychology, 54(1), 15-28.
Leech, K. A., Wheat, D., Rowe, M., Blatt, J., & Dede, C. (2023). “Literacy is everywhere!”: using digital technology to broaden how parents view the home literacy environment. Applied Developmental Science, 27(4), 389-402.
Leech, K. A., Haber, A. S., Jalkh, Y., & Corriveau, K. H. (2020). Embedding scientific explanations into storybooks impacts children’s scientific discourse and learning. Frontiers in Psychology, 11, 1016.
Leech, K. A. & Rowe, M. L. (2021). An intervention to increase conversational turns between parents and young children. Journal of Child Language, 48(2), 399-412.
Lin, G. C., Schoenfeld, I., Thompson, M., Xia, Y., Uz-Bilgin, C., & Leech, K. (2022, June). “What color are the fish’s scales?” Exploring parents’ and children’s natural interactions with a child-friendly virtual agent during storybook reading. In Interaction Design and Children (pp. 185-195).