How should a wide variety of educational activities be sequenced to maximize student learning? Although some experimental studies have addressed this question, educational data mining methods may be able to evaluate a wider range of possibilities and better handle many simultaneous sequencing constraints. We introduce Sequencing Constraint Violation Analysis (SCOVA): a general method for evaluating alternative activity sequences using existing data. SCOVA can be used to explore many complex sequencing constraints, such as prerequisite relationships, blocking, interleaving, and spiraling. We demonstrate SCOVA on data collected from a fractions intelligent tutoring system (ITS). Some of our findings challenge our initial hypotheses regarding sequencing, illustrating the utility and versatility of the method. The method can also be applied to other learning environments, as long as the available data has substantial variability in students' activity sequences. [For the full proceedings, see ED592609.]