A system was envisioned which would enable the graduate-level social science student to use a computer to construct statistical procedures according to theoretically significant formulas, rather than to provide output in the most efficient manner. It was hoped that this system would allow the student to spend more time learning to manipulate statistical data and less time learning to program the computer. Thus far, a number of sub-routines have been programed which provide simulated data and which allow the user to modify this data and to construct standard research statistical techniques. Some work has been done on a control program which will consist of a series of "pre-compiler" statements to be incorporated automatically into student programs. This control program will provide the technical "overhead" programing and relieve the student of the technical concerns. This report stipulates the design criteria and the programing strategy for the system. It describes briefly the procedures already compiled and the present state of the control program. A proposal for future work to complete the system is presented. (JY)