January 24, 2016Strategies in Qualitative Research Conference Creating New Histories of Learning for Math and Science Instruction: Using NVivo and Transana to Manage and Study Large Multimedia Datasets.
What is the Wisconsin Center for Education Research? 37 year old education research center 350 faculty, staff, and students 50 different projects - from single researchers to teams of $17.5 million annual budget 89% federally funded research Technical Services – staff of 14
Growing Importance of Multimedia Data in Research TIMMS-R – over 200 hours of video of math and science classroom practices in each of the 7 participating countries Multiple studies of comparative curriculum redesign – hours of video per year across 4-5 years 15,000 hour of video from a nationwide sample of children from birth to middle school to study the long-term impact of childcare
New Media Tools Emerge From Universities & Research Centers Clan, SALT, Transana, and others from the U.S MediaTagger, Sequence 5, Transcriber, ATLAS.Ti, and C-I-SAID emerge from European research efforts EMU and the QSR tools are developed in Australia
Free Tools Don’t Scale & Commercial Stuff = $/£/¥/€/Etc. Faced with the enormous sizes of many clinical video collections, standalone analytical tools don’t provide adequate collection management support Commercial digital video management solutions cost between $30k-$150k for a system capable of handling 4-5TB of video Lack of standards inhibits interoperability
Transana (and Clan) As Prototype Platforms
Data Management Plays Out As Key Feature in Design
Alternate Operationalizations
Exporting of Media Clips Still Done by Hand
Applying Analytical/Descriptive Coding to Linear Media Video segment as unitary object Interaction depicted by video has multiple phases that have temporal order and length Multiple timelines – along the lines of a video editing tool – may be an answer Small, extracted video segments embedded as data bites is another reasonable approach Using colored shapes drawn on or embedded in media files Multiple transcripts – different methods
Video Editing Tools Provide Possible Coding Interface Synch’d video Multiple transcripts Management window Use(r) rights
Features on the Drawing Board - Based on Current Technologies Data cutter integration Clip representation as URLs Export out to free-standing dataset Export to/import from CHATxml format Server-based, phoneme-based text recognition for collection searching (FastTalk & BBN) Visual object recognition and search (Oracle) DRM & linked use models (METS & TalkBank)
Contact Info Chris Thorn Wisconsin Center for Education Research 1025 W. Johnson St., Room 370 Madison, WI USA