impossible alternative, the instructor may give him a command that makes no systematic contributions to subsequent data values. The Pedagogy: a Simulated Scientific Community The classroom structure of the course is that of a scientific community where each student plays the role of an individual researcher. The student participates in activities designed to parallel those of a scientist in a real community. For example,the student is directed toward a particular problem area. He becomes familiar with at least some of the literature in that area and examines current theories and research bearing on it. He is acquainted with some of the costs associated with conducting research in that field, and he is informed of the resources available to him. These resources, usually in the form of points, are designed to parallel those known to exist in the real world. He proposes research that he would like to conduct, bringing to bear what he is learning about the subject matter, research design principles and his available resources. He designs a research program, not merely isolated experiments. He argues for his research strategy articulating how he believes such a strategy will accomplish his stated COMPUTERS IN MEDICAL RESEARCH Scientists at the Texas Institute for Rehabilitation and Research in Houston are using computer produced three-dimensional measurements to study medical applications ranging from deformities in children to weight loss in astronauts. Dr. R. E. Herron, left, Director of the lnstitute's Biostereometrics Laboratory and Professor Jaime R. Cuzzi, study contour maps of the body which can be produced in other ways such as a graph showing how body volume is distributed from head to foot and cross sections or slices of the body. (Photo IBM) research goals. Upon receiving his "contract", he proceeds to implement his strategy by conducting the experiments he has proposed. He updates his knowledge about the area by re-examining or modifying later experiments, based on what he has learned from his own research and other ”scientists" in his "community". He communicates his updated knowledge to the rest of the "community" in the form of research reports written in a format acceptable for publication, through formal presentation at "conventions" and through informal bull sessions with other members of the community. ln this communication, he not only articulates his research goal, hypotheses tested, and experimental designs, but his method of data analysis and his conclusions. On the basis of updated knowledge, he plans his next experiments, taking into account the costs of such experiments, and the cycle is repeated. One of his final scientific communications may be a review paper that summarizes the state of knowledge of his scientific community in a particular area. It may be a report to a sponsor. Within a two-month period he will have designed and analyzed the results of about ten experiments that he himself has conducted on several problem areas. The simulated scientific community is facilitated by the use of computer data-generating models. The student rehearses all of the major roles of the scientist except the very time-consuming and expensive data collection step. (Even this step can be included in some problems and in some sessions of the course, but the simulation does not depend on its inclusion.) This process is summarized below: Sequence of Student Activities in Simulated Scientific Community 1. Orientation to the Scientific Community a) Activities involved in participating in the scientific community. b) Familiarity and experience with the computer in terms of running EXPER SIM experiments. c) Discussion of issues and problems associated with the particular model in the library that will be used. d) Explanation of simulated costs and budget. 2. Students submit a proposal of their research plan for approval by the instructor. 3. Upon approval of research plan, students submit their first design to the computer and write a short report that will either be posted or dittoed for all class members. If the latter, then it may be viewed as Vol. 1 of a journal for which the students can select a name and formulate editorial policy. 4. Students, now working individually and in collaboration with other students, submit the next design to the computer. lt should be based on what has been learned from the first set of experiments. This report goes into Vol. 2.