New York University

To support the development of and data exchange between Datavyu (a tool for video coding) and Databrary (a platform for archiving and controlled sharing of video data)

  • Amount $550,000
  • City New York, NY
  • Investigator Karen Adolph
  • Year 2017
  • Program Technology
  • Sub-program Data & Computational Research

In the behavioral sciences, fields like child development and behavioral ecology often rely on video as a primary source of research data. The major research repositories of behavioral video, however, do not have much more sophistication than YouTube, relying on keywords and transcripts for discovery and failing to leverage incredibly sophisticated coding data into analytic tools. Karen Adolph at NYU and Rick Gilmore of Penn State are accomplished psychologists who are responsible for developing a leading video coding tool, the open source Datavyu, and an innovative platform for behavioral video archiving, Databrary. The former is notable for its flexibility and fine-grained resolution, and the latter for its ability to set precise access controls to comply with the myriad restrictions related to the use of human subjects of the projects whose data it hosts. This grant supports an 18-month project by Adolph and Gilmore to use Datavyu and Databrary to model integration between coding tools and data repositories more generally. Since both platforms are open source and have active user communities, they are excellent candidates to prototype how standards-compliant coding data might be transferred into a data repository alongside its raw video, and how that repository might then leverage that coding data into new discovery and analytic interfaces. This work could generalize to a host of other coding tools, not to mention the handful of other social science data archives like ICPSR and Dataverse that are tentatively moving into hosting behavioral video data.

Back to grants database
We use cookies to analyze our traffic. Please decide if you are willing to accept cookies from our website.