DevCon: Are large courses bad? A case study in educational data science

File uploaded by katie.desrosiers on Aug 9, 2017
Version 1Show Document
  • View in full screen mode

In education data, some relationships within data are obvious. For instance, more studying improves performance on tests. Some relationships are less obvious, and require data analysis to get an understanding of. This session will demonstrate how to use data to answer these questions in a reliable, scientific manner.

This session will walk through a case study of finding out how course size impacts students’ performance in a course. This process will include data exploration using visualization and statistics, data modeling and machine learning, as well as how to validate findings and decide if they are actionable or not.


Presenters: Derek Kromm Bryan Ohaver