Walkthrough 1: The education data science pipeline with online science class data

This chapter explores tidying and transforming data from online K–12 science courses. Particularly for those carrying out educational research or evaluation, data from online courses and learning management systems (LMS) is commonly analyzed by data scientists in education and those involved in learning analytics. A feature—but also a challenge—of this type of data, which comes from both surveys as well as measures of students’ interactions with course content (i.e., “trace” measures), is that it often requires substantial time and effort before it can be described, visualized, and modeled. This type of analysis can be useful for understanding students’ experiences in online courses; it can also be used as a part of a process of providing feedback to teachers and students, as is commonly the case with learning analytics approaches. Data science tools in this chapter include joining together different datasets, pivoting data from “long” to “wide” form (and vice versa), and exploring data through visualizations, correlations, and regression models.