An optimization-based DSS for student-to-teacher assignment: Classroom heterogeneity and teacher performance measures

摘要

A significant amount of administrator's summer planning is spent attempting to assign hundreds of students to teachers. These assignments must satisfy federal guidelines, parent preferences, and principal preferences while pursuing classroom parity and equity in regards to academic performance, behavioral support, and demographics. In addition, teachers are often the most invested in these decisions given the impact on their day-to-day life and their future performance evaluations. This problem is a variant of a standard problem in operations research; however, administrators often lack the technical expertise or budgets to build and/or implement such models. We present an open-source spreadsheet model-based DSS to this purpose. We formulate the problem as a mixed integer program, present the implemented spreadsheet model interface, and illustrate on a set of student data. This DSS was implemented in the fall of 2018 for the assignment of close to 600 students to 24 classrooms saving dozens of staff and administration hours.With minor adaptions and/or redefinitions, the DSS would provide similar value to many of the roughly 64,000 public primary and middle school principals across the country. Even more generally, the model as presented is a framework for developing and implementing an optimization-based DSS for assistance in creating project teams under a variety of potential preferences and constraints on team composition.