Share This Page
KidTalk: A Natural Experiment in the Time of Covid-19
June 30, 2020
A few years ago, my mother came across an old cassette tape of me singing the ABC song at around two years of age. She recounted how she had made copies of that tape to send to both sets of grandparents, who lived several states away, so they could hear my language progress (and laugh a bit at how I rushed through the L-M-N-O-P part). Today, parents do not need to rely on tape recorders to collect and share such memories. They capture snippets of their children’s songs, utterances, and unique word pronunciations with smart phones, and often share them with friends and family on social media platforms. Now, researchers at Boston College and the University of Maryland have released a new citizen science app, KidTalk, a language scrapbooking tool, to help parents create timelines of their children’s evolving speech development—not only for their own family archives but to provide data to help developmental psychologists better understand language development and how it may be affected by the current Covid-19 pandemic.
Language in the Laboratory
Traditionally, developmental psychologists have conducted their research within the walls of the laboratory. To do controlled studies and examine a single experimental factor, many scientists had to make trade-offs when it came to ecological validity, or how well those results would stand up in a real-world environment, said Yuko Munakata, a developmental neuroscientist at University of California–Davis.
“In the lab, as we try to control so many factors to get a sense of different cognitive processes, we lose a lot of the real world context that shapes behavior, including what may be motivating different behaviors including the way a child speaks,” she said. “We can measure specific capacities in a clean way, but, in doing so, we lose other things that may be very important to how a child is developing that particular capacity.”
For example, Yi Ting Huang, head of the Language and Cognition Lab at the University of Maryland, is interested in better understanding the correlation between socio-economic status (SES) and language development. Many studies have suggested that children who grow up in a lower-SES environment show speech deficits in comparison to higher-SES peers.
“We still don’t know what it is about SES that makes such a difference,” she said. “There are the issues of social and economic adversity. But there are also issues of race, culture, and language dialect, the number of books in the home, how many words a child is exposed to, and other things. There’s a lot to really understand here.”
After playing recordings of her young daughter speaking to her developmental psychology students, she started to wonder if she and other researchers could use the audio from the videos that so many parents take of their young kids to help them get at some of that complexity.
“My students thought I was recording my daughter for teaching or research and I was like, ‘No, I record them for her grandparents!’” Huang said. “So many parents do these recordings now. It occurred to me that smart phones and other devices are high quality enough now that we can do speech analyses. Everything we need in the lab can now be found in most people’s homes. So, if we can encourage parents to participate, we could collect a lot of data right in the home and study the different phenomena we haven’t been able to really investigate before.”
A Natural Experiment
As Huang was discussing this idea with her colleague, Joshua Hartshorne, head of the Language Learning Lab at Boston College, Covid-19 started to spread across the United States. All of a sudden, finding ways to continue research efforts while colleges and universities were closed became a priority. The pandemic, which closed daycares and schools across the country, also raised concerns about how those moves might affect child development.
“Between the lack of school and daycare, the social distancing, the additional parental stress, the lack of school enrichment activities, and a presumed increase in screen time, parents were worried about how the pandemic was going to affect their child’s normal development,” Hartshorne said. “Yet, it’s hard to know what the impact actually is because we can’t test kids in the lab right now. If we could find a way to effectively test things in the home, Covid-19 could offer us a unique opportunity to do an incredible natural experiment.”
Huang agreed and added that Covid-19 could also offer them a new lens with which to examine how changes in SES may affect speech acquisition. “Even high SES families may be experiencing new social and economic stress right now,” she said. “So, we may be able to see the extent to which that stress affects kids independent of typical cultural and race factors.”
To collect this kind of data, Huang and Hartshorne developed KidTalk, a citizen science experiment where parents can create a language scrapbook by uploading their video and audio recordings (video recordings are automatically stripped to only audio information) to an app. There, those recordings come together to form a timeline of speech development. The researchers hope that parents, who are already recording their kids, will appreciating having a “bird’s eye view” of their children’s development as well as a platform to share different recordings with friends and family. The researchers will combine the data from all the participants, keeping each child anonymous, to better understand how Covid-19 may be changing children’s learning environments and overall speech development.
“We’re hoping to get a lot of data from a lot of different people,” said Hartshorne. “Doing it as a scrapbook is a nice way to get that data in a way where parents can record the things they want, go back and listen to what their kid was saying months ago, and share that information with others. I think it’s the kind of experiment that can benefit everyone.”
The app is now available for download online at the KidTalk website. People can start recording immediately from the app and extract audio from videos they already have. Huang and Hartshorne hope to amass a large data set that can also be accessed by other scientists in the future.
“Our research questions aren’t the only research questions that can be answered using this platform,” said Huang. “From the beginning, we’ve been thinking about ways to make this data set larger and more accessible so other labs can use it for their work. Science is something that belongs to everyone – so we hope that others will use this data for their own research.”
Munakata, whose own lab has been considering ways to continue their research while out of the lab, said KidTalk is a creative approach to collecting a rich set of data from a natural setting. She hopes that the results from KidTalk, when coupled with traditional laboratory studies, can provide a much richer view of speech acquisition and development long after the current pandemic ends.
“Many developmental scientists, over the years, have tried new ways to try to capture the richness of development in an ecologically valid way,” she said. “And as more do so, whether because of Covid-19 or for other reasons, it’s my hope that it will drive many more scientists to think beyond the way we’ve always done things to capture those little contextual factors that have the power to influence a child’s development. It provides us with insights we probably wouldn’t be able to get at otherwise.”