EvidenceBase is a relational database system and navigation interface that helps educational researchers to explore and analyze many thousands of text and rich-media data sources. These include hundreds of survey response sets and hundreds of hours of video, all gathered from nearly 300 students and more than a dozen teachers during a four-year-long, school-wide longitudinal investigation of innovative curricular practices. EvidenceBase accomplishes this by leveraging the power of Filemaker Pro -- a highly configurable, cross-platform tool that can be deployed securely and reliably on local machines, read-only media such as CD and DVD ROMs, and local and wide-area networks.

EvidenceBase shows how independent development, close collaboration with team members, and thorough project familiarity can fuel successful development of an innovative tool for educational research.

My Role

I developed EvidenceBase independently, in my capacity as coordinating manager for the project whose information it contains and connects. In doing so, I collaborated closely with my immediate supervisor and project colleagues, and trained them in aspects of its use. Even so, I remain a principal direct user of EvidenceBase, and rely on its export capabilities to turn out Excel spreadsheets, image-form charts, and PDF documents upon request by the project’s principal researchers, that connect across its various kinds of data in highly configurable ways.

The Challenge

Research projects take shape through continual refocusing and refinement. The same is true to an extent with any project - as evident from the PMBOK's five process groups and 13 knowledge areas - but is doubly so for projects where the major outcome is supposed to be new knowledge. Inevitably, a lot of learning happens along the way, and a great many occasions where you run some kind of test and learn enough along the way to merit a refocusing; this is in fact a description of the overall dynamic of project-based research. As the saying goes, “if we knew what we were doing, it wouldn’t be called research.” Move that along a longitudinal axis of multiple class years, and scale it up to an entire school, and you’ve got a tremendously complex data organizing challenge on your hands.

Not too surprisingly, source data was of many different kinds, and subject to many different kinds of analysis. Much of this source data was “born digital,” such as students’ classwork and teachers’ planning texts made available to us as word processor files, interviews recorded on cell phones and other audio-recording devices, classroom interactions captured on digital still cameras and camcorders, spreadsheets of school records and survey responses originating on networked databases, and still-image captures of computer screens. Additional source data was “born analog” and had to be converted to digital forms so it could be included in the project’s dataset; students did most of their classwork and homework using pen and paper, for instance, and a good many school records were available to us only in printed formats.

A Brief User Guide

A complete introduction to EvidenceBase is beyond the scope of this portfolio entry. For this reason, I am appending a document that I had written to a section of the tool that afforded analyses for data from several sets of student essays, and assessments of those essays by expert raters.