Provides a summary of a series of research and development projects which investigated the reasons for failure in educational technology-based systems of instruction. Develops a new approach to the design of such systems and their subsequent implementation and control. A 17-item bibliography is included. (Author/CHC)

Software design is still a craft; techniques are based on the way things have been done, not on basic theorems or design principles. Suggests that academic software design needs to be more like a science or technology and that a taxonomy of design solutions or classification method is required. Describes two case studies. (MLW)