Programmed instruction was defined in the early 1960s as using active response by the learner, immediate confirmation of correct responses, and successive approximations toward the knowledge to be learned. This paper proposes some newer and presumably more effective techniques of programmed instruction and the theoretical justification of those techniques.