Capturing Children's Early Learning Environments

In Collaboration with: Dr. Bria Long, Dr. Michael C. Frank, Dr. Virginia Marchman


BabyView Project

The BabyView Camera is a professionally designed child-safety helmet with an attached Go-Pro designed to capture an egocentric view of children's early learning environments. I am a project manager on a longitudinal study that collects high-resolution, at-home egocentric video data from children 6-30 months over a 2-year period. Using this data we can concurrently analyze the statistics of a child's naturalistic visual and auditory input and track changes in these inputs with the child's developmental trajectory. This rich data allows us to quantify variation in children's early ecology and build models that help us understand how the child's ecology scaffolds early learning.

Design documentation, assembly instructions, and participant instructions can be found at

Long, B., Kachergis, G., Marchman, V. A., Radwan, S. F., Sparks, R. Z., Xiang, V., Zhuang, C., Hsu, O., Newman, B., Yamins, D. L. K., & Frank, M. C. (2023). The BabyView Camera: Designing a New Head-mounted Camera to Capture Children’s Early Social and Visual Environments. Behavior Research Methods.

PreSchool ChildView

One challenge in attempting to quantify children's everyday contexts is capturing naturalistic accounts of ecologically viable contexts outside of the home. As an extension of the BabyView Longitudinal Study, I also manage a project where I collect egocentric video data of children's naturalistic preschool classroom environment for children 3-5 years-old. This data set aims to collect naturalistic accounts of child-led play, exploration, and social interactions in the classroom context. Using this data we can analyze similar statistics of visual and auditory input and also analyze novel first-person accounts of naturalistic child-child and child-teacher interactions.

Ultimately, our goal is to make a publicly available (for authorized researchers) dataset comprised of video data from both BabyView and ChildView Projects.


Inferring Knowledge from Communication

In Collaboration with: Aaron Chuey and Dr. Hyowon Gweon


Identifying Knowledgeable Speakers Using Causal Influence

How can we learn from observing communicative exchanges? Prior work has explored children's understanding of how speakers can influence listeners' beliefs and behaviors. This work asks whether children make knowledge inferences about a speaker using their causal influence over listeners. Across 3 studies, we test whether children are sensitive to changes in the outcomes of a listener's behavior (1), changes in a listener's behavior (2), and whether a speaker spoke or sneezed (3) as evidence for identifying knowledgeable speakers. We find that by 5 years-old, children can reason causally about the consequences of communication in order to infer the knowledge of a speaker. Future work could aim to explore (1) more complex multi-party inferences made in these social contexts and (2) implications for these abilities in pedagogical contexts where knowledge-inferences may support children's ability to assess and learn from knowledgeable informants.

Study materials (videos & stimuli) and data can be found at

Chuey, A., Sparks, R. Z., & Gweon, H. (2023). Young children can identify knowledgeable speakers from their causal influence over listeners. Proceedings of the 45th Annual Conference of the Cognitive Science Society, 230235.
Sparks, R. Z., Chuey, A., & Gweon, H. (2022). Preschool-Aged Children Can Infer What Speakers Know Based on How They Influence Others. In Stanford Digital Repository.