stitutions are able to carry part of the burden of educational computing. As time goes on, these institutions will improve the service they give their undergraduates, while smaller and poorer institutions will be trying to catch up. Many of them will be able to catch up to today's best in 10 or 15 years instead of the 5 years we recommend. If the deficit in educational computing is not made up quickly, millions of students who will have attended these institutions in the 1970's will be poorly prepared for the world of the 1980's and 1990's. The answers we have arrived at are intended to apply over the period from late 1968 to late 1971. The new generation of large computers which is just coming into use will be predominant during most of this period. It will be a period during which techniques and apparatus which are now available and in limited or experimental use will become widely used. Although general computer technology will improve, we believe that during this period there will be no widespread revolutionary effects due to such advances as "microelectronics" and ultra-large memories. The cost of providing adequate computing turns out to be large in overall magnitude, but the estimated cost of $60 per student per year averaged over all college students is comparable to the $50 to $200 per student per year for college libraries and an estimated $95 per chemistry student per year for a single chemistry laboratory course in a 4-year college. The total cost for adequate computing in connection with training in computer sciences will be a small part of the total cost of adequate computing for undergraduate education, so the average of $60 per year should provide for adequate educational use of computers in colleges and universities (exclusive of graduate research). While the average cost of $60 per student per year is a small part (around 4 percent) of the overall educational cost per student per year, there is no place for it in the already tight budgets of America's colleges and universities. Further, the cost is growing rapidly compared with other expenses, and it must grow even more rapidly if adequate service is to be provided for all students. We believe that it is in the national interest to have adequate computing for educational use in all our institutions of higher education by 1971-72. We believe that this can be achieved, but we believe that it can be done only with government assistance. The cost and use of computing in colleges and universities will and should be a small part of the total cost and use of computers in industry and in government.* But these major economically and socially productive uses are dependent on educational computing not only for the training of manpower, but for stimulation both in new and productive uses of computing, *In 1965 the capital value of college and university computers was $300 million, or one twenty-sixth of the total of $7.8 billion for the United States (fig. 4 of App. I) and the cost of computers used in instruction $35 million, or one-two hundred twentieth of the U.S. total, according to the report "Digital Computer Needs in Universities and Colleges,” (Rosser report), publication 1233, National Academy of Sciences. and in advances in computers and their software. Adequate support of computing as a part of education is essential for a rapid and full realization of the social and economic benefits of computing. While an investigation of the cost of and means for remedying our computer deficit in undergraduate education by 1971 is a chief purpose of this report, other purposes are to cast light on various opportunities and problems in the educational use of computers. The report is divided into four main sections. The first is primarily concerned with computers and undergraduate education. The second considers education in computer sciences, mainly at the graduate level. The third discusses the interaction between research and educational use of computers. The fourth comments briefly on computers and secondary education. The major findings and recommendations of the Panel are: 1. Approximately 35 percent of college undergraduates are enrolled in curricula in which they could make valuable use of computers in a substantial fraction of their courses. An additional 40 percent are in curricula for which introductory computing training would be very useful, and limited computer use should be part of several courses. The remaining 25 percent could make some use of computers in one or more courses during their college education, but computer training is not now important in their major studies. In 1965 less than 5 percent of the total college enrollment, all located at a relatively few favored schools, had access to computing service adequate for these educational needs. However, it is practical to supply adequate computing service to nearly all colleges by around 1971–72. We recommend that colleges and universities in cooperation with the Federal Government take steps to provide all students needing such facilities with computing service at least comparable in quality to that now available at the more pioneering schools. 2. One of the major problems in providing the necessary educational computing is the cost. The yearly cost of providing this service will rise to a total (for baccalaureate programs and 2-year colleges) of about $400 million per year in 1971-72 in addition to the relatively smaller costs required for faculty training and associated research. It is beyond the capabilities of our colleges and universities to bear all of this cost in this time period. We recommend that colleges be encouraged to provide adequate computing through government sharing of the cost. Such governmental cost sharing should include special grants to cover transient costs when service is being initiated or larger facilities are being inIt should also provide a portion of the annual cost of continuing service. stalled. 3. Government accounting practices have made it very difficult for colleges and universities to utilize fully that Federal and private support for computers or computer service intended for unsponsored research and education (as distinguished from research paid for by grants and contracts). Treatment of a grant for educational use of a computer as a reduction in total cost reduces the hourly charge for computer time paid by all users and has the effect of shifting research costs to educational users. The Department of Defense has recognized this and now has an agreement with the National Science Foundation not to treat NSF educational grants for operating expenses as a reduction in sponsored research costs. Many schools cannot now afford to pay for educational and unsponsored research use of computers by students and faculty even though there is time available on their computers. Consequently, some college and university computers now available for educational and unsponsored research use are standing idle for major portions of the operating week. We recommend that the present DOD-NSF agreement be extended to other government agencies and private supporters and include both capital and operating cost grants. Additional Federal funds should be made available immediately for support of computing service used for education and unsponsored research activities at institutions presently having the required facilities. 4. We find that any expansion of the educational use of computing depends heavily on increased knowledge of computing by faculty in most disciplines. Such knowledge usually can be provided by intensive 2- to 6-week periods of faculty education. The extensive activity of the National Science Foundation in sponsoring summer institutes provides a useful model. We recommend an expanded faculty training program to provide adequate faculty competence in the use of computing in various disciplines. 5. There is a great need for specialists trained in the computer sciences at the bachelor's, master's, and doctorate level. The whole success of educational computing and continued improvement in its use depends on expanded education and research in computer sciences. This education requires a good faculty and access to very good computing facilities for both course work and research. We recommend that the Federal Government expand its support of both research and education in computer sciences. 6. The cost of computing is a continuing expense, like light or water, rather than a capital investment, like the initial cost of buildings. We recommend that the Government agencies which support computing allow the schools to be free to apply the funds either to the purchase or rental of equipment and the support of staff, or to the purchase of service. 7. The optimum mechanism for providing computers will differ from campus to campus. However, in many cases it appears economical and ef fective to supply adequate and dependable service from large computing centers. We recommend that universities and the Government cooperate in the immediate establishment of large central educational computing facilities capable of serving several institutions. 8. Because of inconsistent Government and university accounting practices, the great variety of sources of computing support, and the experimental nature of computer use, some universities have had difficulty in determining and controlling their computer costs. Informed decisions regarding expansion and/or budgeting for current operations cannot be made without accurate cost information. Errors made at this stage can only lead to the diversion and dissipation of university resources needed for other educational purposes. We recommend that universities and colleges develop and use accounting procedures which accurately measure the cost and utilization of computer services. With such information the allocation of computer time for research and education and the anticipation of associated costs should be made on a realistic and measurable basis. 9. Proper introduction of computing into secondary education is desirable and growing. Not enough is known about the best ways for introducing computing and we were not able to consider this adequately in the time available. We recommend that NSF and the Office of Education jointly establish a group which is competent to investigate the use of computers in secondary schools and to give the schools access to past and present experience. Cooperation between secondary schools and universities, and particularly providing service to secondary schools from university centers, should be encouraged. 10. There is inadequate information about the number and level of skills of personnel now employed in the field of computers, and there are no meaningful forecasts. We recommend that the Federal Government collect meaningful data concerning computers and the jobs, personnel, and educational facilities associated with them, and endeavor to make useful annual forecasts. II. COMPUTERS AND UNDERGRADUATE EDUCATION Computers in Colleges and Universities Computers were first introduced into universities as rare and special pieces of equipment used for a few specialized sorts of research by small groups of people. Today, many universities and colleges have centers which serve most of the students, faculty, and administration both by providing training in programing and by meeting computing needs for undergraduate education, for research, and often for administration. Where adequate computing facilities have been available, the faculty has made increasing use of computing in both research and education, and computing has become a part of more and more undergraduate courses, including business subjects, social sciences, biological and health sciences, psychology, geology and other disciplines, as well as mathematics, physics, chemistry, and engineering.* This is consistent with the rapidly growing use of computing outside the schools in small as well as large business enterprises, in government operations and national defense facilities, and in almost all technology—those many fields of endeavor where most college graduates will find their places. Computing is not an esoteric or specialized activity; it is a versatile tool useful in any work with a factual or intellectual content. Computing is becoming almost as much a part of our working life as doing arithmetic or driving a car.** Computers find a widespread use in education only when well-run facilities are easily available to all students and faculty members, with rapid service for all users. Under these conditions there are a number of instances (including, for example, Dartmouth and Texas A. & M.) in which a majority of all undergraduates learn programing and use computing in some part of their course work. While computing has not yet become an important part of undergraduate course work in such fields as English, linguistics, languages, history, music, and art, faculty members in some of these fields are making increasing use of computers in research, and computing is beginning to find its way into undergraduate instruction. In all fields where computing has been used, it has added a new dimension to education, and has led the students to better comprehension of complex *The examples given in App. J give some idea of the range of profitable use of computing in higher education. **Apps. B, C, and E introduce some of the jargon, e.g., hardware, software and |