Asking questions about data: First-year engineering students' introduction to data analytics

Abstract

This complete research paper aims to understand the question design's process of first-year engineering students when performing data analytics. Specifically, we aim to answer the research question: How do first-year engineering students use a large data set to ask questions focused on the client's needs? While most research in the area of analytics has focused on how to perform the data analytics cycle successfully, the learning process behind the practice of data analytics is still not totally understood. This study was conducted with 53 first-year engineering students who worked in 14 teams to solve an open-ended problem of data analytics called: The Bikeshare Problem. Students were tasked to download the freely available data from Capital Bikeshare company (~3 million data points) and do a preliminary analysis to understand the data set and the company itself. Students proposed questions individually to explore the data and, as a team, design a team question focused on the client's needs. We used content analysis to develop a codebook and analyze the 92 individual and 13 team questions. Our results showed students were able to take the data and frame a problem until designing a question that considered the client's perspective. However, some students ended up writing questions that were not clear due to ambiguous terms, so simple as not to be useful to the client, or too complex to be answerable with the date and time they had. Additionally, students' questions often focused on a single, simple variable of the data and not utilizing the breadth of the data available to them. Finally, the student's previous knowledge of statistics could have mediated their question design practice and limited their ability to answer their questions. We presented some suggestions for researchers and professors who want to study and teach analytics. The Bike-share problem is an example of how analytics can be successfully integrated early in engineering curricula, and we animate professors to implement similar activities in their courses.