Video-based Research and Professional Development Project (VRPD)

Project Director: Ken Hay

The VRPD project has as its main mission the creation of a video capture and analysis system, providing an infrastructure that will allow learning scientists and students in training to generate and advance warranted assertions about the learning and cognition. Specifically, we would like to prioritize support for developing and validating methodologies and video based technologies that enable the collection of pervasive, systematic, and direct evidence of learning and cognition as they unfold in naturalistic contexts. Furthermore create powerful means to bridge the chasm between research and practice through new tools for web-based video research publications that directly target practitioners. Over the next two years, we will build on existing infrastructure developed through funded work to establish a general video capture system that will emphasize a range of different video capture needs from single camera capture of a presentation to a multi-stream capture involving group work and 1-to-1 instructional computing. We will continue to develop data archiving systems and strategies that will support a wide array of users, while maintaining data security and possibilities for collaborative analysis and supporting research computing support services to effectively support researchers at IU and other universities.

Our vision is a system that supports multiple analysis of video and leverages our data archiving systems to make video access dramatically reduce initial biases. Source data can be indexed and "viewed" through different, complementary or competing perspectives without compromising the source data. Our data display interfaces will thread different analysis together to form a blended representation. This allows researchers to challenge or build upon different analyses of the same event. Furthermore it will support collaborative video analysis that will promote geographically distributed research collaborations. This will be possible because aggregation-level metadata (i.e. data about all TMD involved with a particular group at a particular time) will be used to dynamically construct web interfaces that flexibly blend the data streams. The system will also support code development, instrument validation, collaborative video interpretation, visualization sharing, and tentative hypothesis testing. Collectively, this creates a potential for intensely iterative research across distributed research communities.

The reporting of video-based research via textual transcript passages in written articles often leaves the potential for research impacting the community wanting. It also leads to criticisms that the author is simply "cherry-picking" the best, and perhaps the few, examples. We will build a system that supports that distillation of video data into web-based or CD-based publishable formats that will support the power of video to exemplify our research. Furthermore, it will enable authors to not only select "an" example, but connect multiple examples of the illustration to present and make available to the reader.

The Vice President of the American Council of Learned Councils has criticized social science research as having a "context of disconnection" between research and practice (in Willinsky, 2001). The groundbreaking potential of primary multimedia data to bridge this void is highlighted by several projects recently funded by the NSF (e.g., Carraher & Nemirovsky, 2001; Miller, et.al., 2000; Siegler, Gallimore, & Hiebert 2000) where leading-edge projects use digitized video to disseminate and publicize findings at the conclusion of their efforts. The system will support the ability to publish research findings in formats that are appropriate for professional development. This could include formats for presentation, but also formative testing and application of ideas to video based scenarios.

Finally, we will design this entire system so that it can easily be used by teachers and teacher educators for video research within a reflective practitioner mode. Where reflections would no longer be based on "recollection" of what happens in the classroom, but rather on the "direct evidence" on what went captured by video. This could transform student teaching and teacher certification.