Research Question: How do families in a variety of economic, social, and demographic circumstances support their children’s early learning?

Data: Our study features a highly innovative “video ethnography.”  Unobtrusive technologies in the form of video cameras placed strategically in up to four rooms in participants’ homes (rooms where most interactions occur) and activated continuously throughout the day and evening for two weeks constitute the primary means of data collection.  Video data were supplemented by survey and interview data collected from families at six additional points of contact that stretched over 3-4 months.

Methods: All of the video clips have been transferred into Citadel – the University’s Secure Research Infrastructure – where they can be viewed but not downloaded. In a pilot project, researchers are accessing NJFS data remotely via Citadel. We are creating metadata that permit users to query a searchable database and identify clips that match particular criteria around household characteristics, room views, day and time periods, participants in the videos, and activities and behaviors. In an exciting interdisciplinary development, we are sharing NJFS video data with two computer vision courses in Princeton’s Computer Science department.

Challenges: The major challenge is curating a video dataset with approximately 460,000 discrete video clips. This much information is impossible to code manually and offers an opportunity to collaborate with Princeton computer vision colleagues to help automate the coding.

Findings: Our efforts are still focused on creating a user-friendly dataset for a worldwide research community.  However, we expect that the videos will reveal unanticipated strengths in lower-income families as well as provide a new appreciation for the daily challenges and struggles facing many families with young children.