Principal Investigator: A/P Tan Seng Chee
, Deputy Director, Centre for Research and Development in Learning (CRADLE@NTU)
Funding Agency: CRADLE@NTU Start-up Grant
To say that NTU is awash with student-produced data is an understatement. Students access course materials to answer questions in either verbal or written format. They respond to specific items or write essays on open-ended enquiries. They participate in clubs and activities, and end up asking new questions themselves. Much of these information are digitally captured by the University’s existing infrastructure. However, how best to allow faculty, staff, and researchers to ask questions about this pool of student-produced data is an open question.
Researchers at NTU’s Centre for Research and Development in Learning (CRADLE@NTU) are developing a platform so interested parties can draw answers from students’ data while maintaining their privacy.
The platform will make a suite of learning measures and relevant student information available in an aggregated form. For instance, if a set of instructors wished to know if performance in a previous course was predictive of performance in their courses, they could query the platform and see the results. If they wanted to know if students who have tutorial sessions later in the week performed better on homework, they could find an answer through the platform. However, if they wanted to know which student was predicted to perform best during the upcoming semester, the platform would withhold an answer.
This balance between maintaining the privacy of data contributors and the value of their data has been a universal research subject worldwide. Too much transforming and obscuring data to protect individuals can skew one’s findings. At the same time, too little obfuscation leaves open the possibility that individuals can be deliberately identified. The project seeks out this balance while still empowering instructors to reflect and take action upon their own teaching and learning practices.