Instructional Design for Teaching Relational Query Optimization to Undergraduates

Abstract

This paper discusses learning outcomes, instructional materials, software tools, and experiences in teaching relational query optimization in an undergraduate database systems course. The pre-requisite topics for the learning module are basic familiarity with SQL query writing, data structures (e.g., search trees and algorithms for linear and binary searching), and the mathematical definition of a relation. The module is designed to fit within a 6-week period with two 80-minute class sessions per week. Quizzes are used to reinforce learning outcomes, and the module culminates with a multi-phase project covering computation of database statistics, logical optimization using algebraic transformations, and performance of selection and join algorithms as well as computing the cost of execution trees. The instructional design poses several challenges: how to define an appropriate scope for the material given the duration and the level of student preparation, how to scaffold learning complex, inter-twined topics, and how to create effective assessments of student learning. One possible module design is illustrated here, along with the lessons learned from deploying it.