On the approximation of utility functions for faculty teaching assignments

Abstract A number of mathematical programming models have been proposed for generating faculty teaching assignments. Usually these models attempt to maximize the sum of the faculty members' utilities for the courses in the assignments. This paper demonstrates that the use of course utilities ignores important interaction effects in the determination of a faculty member's utility for a teaching schedule. As a result, solutions generated by the models can be far from optimal if the true objective of the scheduler is to maximize faculty utilities for their teaching schedules.