Data-First Learning In Statistics
Grant program: Teaching and Learning Development Grant (TLDG)
Grant recipient: Luke Bornn, Department of Statistics and Actuarial Science
Project Team: Jack Davis, Elpedia Arthur Junior, Jacob Mortensen, and Steven Wu, research assistants, and Daria Ahrensmeier, Educational Consultant, TLC
Timeframe: April 2016 to May 2017
Funding: $5,000
Course addressed: STAT 440 – Learning from Big Data
Description: The skills required to work as a statistician/data scientist in modern industry are at a disconnect with our current teaching methods. In particular, statistics courses are often taught in a methods-first approach, with data only entering the picture to support the teaching of methods. In contrast, in industry practitioners are faced with complex, real-world data alongside a business problem, and it is up to the practitioner to select the appropriate method or model. The goal of this project is to build a new course, “Learning from Big Data,” and to study how well it works for the students as well as the instructor. The new course inverts the traditional statistics learning model; by working directly with real-world datasets sourced from open sources and industry collaborations, students will build the skills to aid them in entering the workforce after graduation.
Questions addressed:
- Does the course design work on a day to day basis, from a practical point of view? For example, are students able to access and manipulate data? Are there any significant technical obstacles to overcome?
- Do the acquired data sets serve their intended purpose – to challenge students with problems akin to those seen in the real world and to guide them in achieving the intended technical skills?
- Does the competition/ranking system work as a teaching method?
- Are students obtaining the desired interpersonal workforce skills?
Knowledge sharing: The objective is to disseminate this teaching strategy to other statistics courses at ¶ˇĎăÔ°AV, as well as to the broader statistics community. This will be accomplished through seminars, publications in statistical education journals, and guest lectures at neighbouring colleges and universities.
Keywords: real-world data; statistical learning