Sorry, you need to enable JavaScript to visit this website.

P3-01: Learning Data Science with the Help of a Data Exploration Tool


By Philipp Burckhardt, Christopher Genovese, & Rebecca Nugent (Carnegie Mellon University)


Abstract

"IBM has predicted that the number of jobs for data professionals in the US will reach 2,720,000 by 2020, an increase of 28% from today. With more and more data getting collected, equipping students with the skills necessary to enter a dialog with data is tantamount. For this purpose, we have built the ISLE e-learning framework (Interactive Statistics Learning Environment) and deployed it in various courses at CMU. In our data exploration tools, students can explore and analyze data sets using a simple point-and-click interface, with only the methods that have discussed in class up to that point being available to them. Students can not only analyze data but also write statistical reports by means of an integrated Markdown editor, into which summary statistics and plots can be inserted via a simple drag-and-drop operation. Since the framework contains peer-to-peer messaging functionality, students can form ephemeral groups targeting a specific subject and thereby assist each other in the learning process. In a blended learning context, the input of each individual can be projected on a shared screen and discussed among the class. This allows instructors to treat the group as a whole, and the group itself is empowered to reflect upon its choices. But not only students are enabled to reflect upon their choices, instructors gain access to a virtual research laboratory that allows them to replay each step of a student’s data analysis process: What plots they have generated, what text chunks they have written, and what results they have obtained."


Recording