In this paper we describe the criteria of "Technology I, II, and III," which some instructional theorists have proposed to describe the differences between a formulaic and a reflective approach to solving educational problems. In a recent study, we applied these criteria to find evidence of a "technological gravity" that pulls practitioners away from reflective practices into a more reductive approach. We compared published reports of an innovative instructional theory, problem-based learning, to the goals of the theory as it was originally defined. We found three reasons for technological gravity, as well as three approaches some practitioners have used to avoid this gravity. We recommend that instructional technologists adopt our three approaches, as well as the criteria of Technology III, so they may better develop instruction of a quality consistent with the innovative instructional principles they claim, and that best characterizes the goals they have for their practice.