Self-paced, instructor-assisted approach to teaching SQL

We present a novel approach to teaching Structured Query Language (SQL), which is suitable for both college classroom environment and asynchronous remote instruction. Instead of sitting passively and listening to a lecture, students work at their own pace through bite-sized tutorials, examples, exercises, practical tasks, and quizzes. Their work is checked in real time by an AI-based software platform which also provides instant personalized adaptive guidance. Students must prove mastery of each concept before being allowed to tackle the next one. In this way, they are active 100% of the time, and get much more hands-on practice than in traditional instruction. The instructor does not lecture, which allows him or her to interact with students individually. It turns out that students not only enjoy the one-to-one interaction with their instructor much more than listening to lectures, but they also greatly benefit from it. In this paper we provide a concise overview of the teaching method, and then we focus on automated server-side analysis and grading of SQL queries, which is the cornerstone of the self-paced SQL course. We introduce a number of Python-based SQL analyzers for various types of queries, and present links to three live SQL assignments for the reader to experiment with. Our SQLgrader library is available on Github under an open source license.