Improving Data Science Teaching

Data Fairs, Matchmaking and Collaboration Patterns for Improving Data Science Teaching

School: Edinburgh College of Art

Team Members: Dave Murray-Rust, Benjamin Bach, Anouk Lang, Catherine Lai

Abstract

Data Science is a core development area across research and teaching in Edinburgh: the CDT in Data Science, the Data Driven Innovation agenda in the Edinburgh City Deal etc. A key challenge of teaching data science is working on real data rather than samples curated for teaching. Live data and motivated data holders expose students to the challenges and peculiarities of messy data, while providing opportunities for engagement and motivation as the results of data analysis are valued beyond the classroom.

This innovation project explores ways to connect students learning data science with staff in need of data analysis, enhancing student experience by offering opportunities to work on real world data as part of their education. We will run data fairs , allowing staff to come together and present their datasets to students who need projects, and create a platform for sharing and matchmaking staff-student data science projects. We will guide staff in crafting data briefs that help students to engage with their data, and use these as the basis of matchmaking. We will research the collaboration patterns that are effective in creating good student experience and high quality projects, as well as distilling case studies and best practices for student data science projects.

Our approach will be co-designed with students and data-holders, and evaluated through pre and post questionnaires and the impact of the collaborations on research. As the project grows, it will serve as a nucleation point for staff-student collaborations on data science projects across the University.

Final Project Report

Download the final project report now (PDF)