Scientific research that is conducted without the structure of a supporting scientific paradigm should be labeled pseudoscience in that such research is deceptive or false science. It is argued that much of the research in educational technology is pseudoscience, with the focus on learner control research. Learner control is the design feature of computer-based instruction that enables learners to choose freely the path, rate, content, and nature of feedback in instruction. It is contrasted with program control. Much of the research into learner control constitutes pseudoscience in that it fails to meet the major theoretical and methodological assumptions underlying accepted research methodologies within the positivist, quantitative paradigm. In the first place, the concept of learner control is poorly defined. In many studies the instructional treatments used are too brief to provide learners with sufficient experience for learner control variables to be actualized. In many media replication studies there is a lack of consequential or relevant outcome measures. Small sample sizes and large attrition rates are other problems that plague learner control research. Suggestions for improving the quality of learner control research, including the exploration of alternative research paradigms, are presented. One figure illustrates the continuum of learner control. (Contains 41 references.) (SLD)