Statistical Training and Research: The University of North Carolina SystemAuthor(s): E. Shepley Nourse, Bernard G. Greenberg, Gertrude M. Cox, David D. Mason, JamesE. Grizzle, Norman L. Johnson, Lyle V. Jones, John Monroe, Gordon D. Simons, Jr.Source: International Statistical Review / Revue Internationale de Statistique, Vol. 46, No. 2(Aug., 1978), pp. 171-207Published by: International Statistical Institute (ISI)Stable URL: http://www.jstor.org/stable/1402812Accessed: 01/06/2010 14:44

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=isi.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

International Statistical Institute (ISI) is collaborating with JSTOR to digitize, preserve and extend access toInternational Statistical Review / Revue Internationale de Statistique.

http://www.jstor.org

International Statistical Review, 46 (1978) 171-207 Longman Group Limited/Printed in Great Britain

Statistical Training and Research: The University of North Carolina System

Authors This paper was collated by E. Shepley Nourse, Publications Consultant, Health Sciences, University of North Carolina at Chapel Hill, for Bernard G. Greenberg, Kenan Professor of Biostatistics and Dean, School of Public Health, University of North Carolina at Chapel Hill, who chaired the group which provided information for this history. Participants included University of North Carolina people (a) from North Carolina State University at Raleigh: Gertrude M. Cox, Professor Emeritus of Statistics; and David D. Mason, Professor of Statistics and Department Head; and (b) from Chapel Hill: James E. Grizzle, Professor of Biostatistics and Department Chairman; Norman L. Johnson, Professor of Statistics and former Department Chairman; Lyle V. Jones, Alumni Distinguished Professor of Psychology, Vice Chancellor, Dean of the Graduate School; John Monroe, former Director, Survey Operations Unit; and Gordon D. Simons, Jr., Professor of Statistics and Department Chairman.

Table of Contents

Early Developments at Raleigh 173 Establishment of North Carolina's First Academic Department of Statistics 173 Establishment of the Institute of Statistics 174

Expansion of the Institute of Statistics 175 Department of (Experimental) Statistics 178

Developments of the 1950s 178 Developments since 1960 179

Summary Highlights of Consulting and Research 181

Department of (Mathematical) Statistics 182

Developments up to the Mid-Sixties 182 Developments since the Mid-Sixties 184 Students in the Department of Statistics 185

Department of Biostatistics 186 Early Developments 186 Developments after 1960 188

Other Ventures 190 Notes and References 192

The development of statistics as an academic discipline in the University of North Carolina

System had its origins before Pearl Harbor. Initial growth, in some ways curtailed and in other ways stimulated during the World War II years, did result in a sound base for the acceleration that occurred postwar and the maintenance of high quality and responsiveness to the needs of the state, region, and nation that has characterized developments to the present time.

What is now the Department of Statistics, School of Physical and Mathematical Sciences, North Carolina State University, Raleigh, was established in late 1940. This pioneer depart- ment experienced a rapidly increasing demand for training, research collaboration, and consulting assistance locally and nationally. To help meet this need, two other statistics departments were established at the University of North Carolina, Chapel Hill: in 1946, what is now the Department of Statistics in the School of Arts and Sciences, and in 1949, the

172

Table 1

Highlights of the first fifteen years in the development of statistics in the University of North Carolina System

University of North Carolina at North Carolina Chapel Hill, Departments of: State College at Raleigh, Depart- Biostatistics

Time Institute of ment of Experi- Mathematical (School of Public Other period Statistics mental Statistics1 Statistics Health) developments

1940-41 Exploration by President Frank P. Graham and establishment of department (Gertrude M. Cox, head) in School of Agriculture; start-up of courses, conferences, research, and faculty growth.

1942-44 Establishment of World War II Institute (at impact. North Carolina State, Gertrude M. Cox, director); grant support from General Education Board (Rockefeller Foundation).

1945-46 Expansion to All- Expansion and Establishment of Biometrics Bulletin University status; rapid growth; department started (later regional respon- addition of (Harold Hotelling, became sibility in South William G. chairman). Biometrics, through work Cochran and Journal of the conferences, other outstanding Biometric summer sessions, faculty. Society). professional assistance.

1947-49 Further progress; Jackson A. Rigney Graduate Establishment of Gertrude M. Cox, became head; program department full time to graduate program expanded. (Bernard G. director post; expanded; Greenberg, Institute cited by Quantitative chairman). new Southern Genetics Program Regional started. Education Board as example of regional cooperation.

1950-54 Continued its Increased George E. Expansion; first At Chapel Hill, leadership role. provision of Nicholson, Jr., degrees awarded. Survey Operations

statistical service chairman; name Unit and to University and change to Psychometric region. Department of Laboratory

Statistics. started.

1 In 1965 there was a name change to North Carolina State University at Raleigh. In 1970, the Department of Experimental Statistics became the Department of Statistics; a comparable name change occurred earlier with the Department of Mathematical Statistics, as shown above.

173

Department of Biostatistics in the School of Public Health. All the early developments in statistics had the strong support of Frank Porter Graham, President of the University of North Carolina System, which then included three constituent institutions at Chapel Hill, Raleigh, and Greensboro.

The entity known as the Institute of Statistics, was established at Raleigh in 1944 and was given University of North Carolina System status in 1946. The Institute of Statistics had an emerging leadership role in the field of statistics, especially in North Carolina and in the Southeast. The unique story of the Institute, as told by Frank Porter Graham, featured 'the cooperation of many persons and agencies, and the adventurous spirit of the preeminent leaders in this field'. It was 'an adventure in creative cooperation at one center for training ... urgently needed specialists and leaders' (Graham, 1948), a reputation that has facilitated a focus that continues to be a distinctive feature of international interest.

The following pages include brief historical overviews and comments on distinctive characteristics, educational offerings, research emphasis, and consulting activities of the Institute of Statistics and of each department. Table 1 summarizes the historical context of the first 15 years. Other exhibit materials include lists of Ph.D. recipients through the 1975-76 academic year and present and former faculty.

Outstanding students and faculty, broad balance of theory and application, and the co- operative focus referred to above are recurrent themes throughout this story.

Early Developments at Raleigh Establishment of North Carolina's First Academic Department of Statistics On a train trip early in 1940, President Frank Porter Graham quite by chance met W.F. Callander, United States Department of Agriculture, who expressed a desire to help establish another center similar to the existing one at Ames, Iowa, where the training of statisticians and cooperative research with federal agencies could be done. President Graham told him, 'We will do it at North Carolina State College,' and shortly thereafter the initial developments were under way.

W.F. Callander and C.F. Sarle, United States Bureau of Agricultural Economics, A.E. Brandt of the United States Soil Conservation Service, and others were contacted regarding the type of program and its leadership. Professor George W. Snedecor was asked to suggest names, and in a letter dated 7 September 1940, he recommended five young men, half-heartedly adding, 'If you would consider a woman, I know of no one better qualified than Gertrude M. Cox'. An offer dated 24 September 1940, was received by Miss Cox and she reported for work 1 November 1940, the first woman professor on the faculty of North Carolina State College. The Department of Experimental Statistics, in the School of Agriculture, was formally approved 22 January 1941, by the All-University Board of Trustees and Professor Cox was confirmed as head of the department.

The major objectives of the Department of Experimental Statistics were to provide statistical

consulting, computing assistance, and service courses for the research staff of the North Carolina Agricultural Experiment Station and the School of Agriculture. A series of elementary courses in statistics was developed immediately.

During the first summer, in 1941, basic applied and theory courses were given in statistics and related fields. The teachers were George W. Snedecor, Harold Hotelling, and Gertrude M. Cox, assisted by Robert J. Monroe and Carl F. Kossack. England's Scotland Yard would not permit R.A. Fisher to come to teach, probably because of his active training of foreign students. There were 83 regular students, who now represent a 'who's who' list of statisticians who have developed many of the statistical programs in the United States.

During this same period, in addition to the regular credit courses, three one-week conferences 46/2-D

174

were held for invited research investigators, especially from the Southern states. These conferences (agricultural economics and rural sociology, biology and nutrition, and agronomy and horticulture) were held for the purpose of aiding research workers with some of their more pressing experimental problems; 243 persons registered. These were the first three of a series of conferences to be held in the South.

Robert J. Monroe joined the faculty in 1941, as did Jackson A. Rigney, initially part time with Experimental Statistics and Agronomy. Richard L. Anderson joined the Department of Mathematics at North Carolina State College in June 1941, then became a member of the Statistics group 1 January 1942. Jay Wakeley arrived 13 September 1941.

A joint arrangement by the Department of Experimental Statistics with the United States Department of Agriculture Division of Agricultural Statistics provided resident collaborators Alva L. Finkner and Walter A. Hendricks. The first formal research project between these two groups, including the Department of Agronomy, was entitled, 'Meteorological-Soil- Plant Relationships'; David D. Mason was employed as a half-time graduate research assistant to work on this project.

World War II military service affected many of the people who were associated with the new Department of Experimental Statistics, and developments during this period were necessarily slower than desired. Soon Robert J. Monroe was stationed at Camp Davis, Dave Mason and Alva L. Finkner were in Australia, Harold 'Cotton' Robinson was in Africa, and Jay Wakeley was in the Navy. Marvin Clay saluted the department on his way overseas by circling his Air Force plane low over 1911 Building, where Statistics was then located. In later war years, Richard L. Anderson was on leave at Princeton University, Office of Scientific Research and Development.

The staff members remaining with the department were partially involved with the war effort. Any professor who had had advanced mathematics courses was drafted to help teach mathematics to soldiers sent to North Carolina State College for training. In the fall of 1943 Miss Cox, with the help of Richard L. Anderson, taught two preflight classes in spherical trigonometry. For awhile beginning 1 July 1944, J.M. Clarkson from the Department of Mathematics helped with teaching courses in mathematical statistics. Ralph E. Comstock joined the Statistics staff on 15 August 1943, and at first was half time with the Department of Animal Science. His main research interest was quantitative genetics.

During these years there was a steady increase in the demands for teaching and consulting work. Although the staff devoted most of their energies to collaborating on research investi- gations, they also did fundamental research in statistical methodology and introduced into the curriculum courses in applied and mathematical statistics.

Establishment of the Institute of Statistics

Requests for assistance increased rapidly, and early in 1944 it became clear that substantial expansion of the Statistics program was needed. Requests for permission to organize an Institute and to secure substantial funds to enable further development of the program and to handle contract projects were made to the Director, Agricultural Experiment Station, and directly to University President Frank Porter Graham. He phoned former North Carolina Governor O. Max Gardner, who was serving in Washington, D.C., and insisted that he had to be present at the meeting of the Executive Committee of the Board of Trustees. Mr Gardner rearranged a meeting of the War Mobilization and Reconversion Board, of which he was chairman, and came by train overnight to Raleigh to the meeting on 18 September 1944, when the committee approved establishment of the Institute of Statistics at North Carolina State College and appointed Miss Cox as director.

A proposal was prepared for submission to the General Education Board (Rockefeller Foundation), which President Graham discussed with A.R. Mann of the General Education

175

Board. Frank Porter Graham wrote later: 'The idea that such a fully integrated program requiring the services of the best minds in this field was to be undertaken in the South, was considered in some parts of America and the world to be reckless and even fantastic. It was considered by some to be wiser for a great Foundation to back such an unprecedented enter- prise at one of the highly endowed universities already assured of its own financial stability' (Graham, 1948).

The General Education Board decided that its support would go to the University which had made the proposal and was willing to venture most for its fulfillment. On 7 December 1944, $87,000 was awarded to the Institute of Statistics to provide for expansion of its program.

Early in the plans for the Institute of Statistics, three purposes were established as goals toward which efforts were to be directed: (a) to provide a teaching, research, consulting, and service center for statistical work; (b) to promote the use of efficient statistical techniques in diverse fields of application; and (c) to advance statistics through the discovery of new techniques by theoretical investigations.

It was recognized that, in order to carry on this program, the Institute of Statistics faculty should consist of both theoretical and applied statisticians: the theoretical statistician to have a high degree of mathematical ingenuity and skill, along with the ability to develop new concepts, and the applied statistician to have advanced training in two research fields, in statistics and in a special field of application. Skill for planning surveys and experiments appropriate for many special situations, a consulting ability achieved only through experience knowledge of the subject matter to which the statistical techniques are applied, and ability and willingness to cooperate were considered important.

The work of the Institute of Statistics staff was broadened to include research in math- ematical, physical, and industrial statistics; before this time, consultant assistance had been provided primarily in the biological and social sciences. The establishment of the Institute with additional funds made possible the development of a more extensive program of co- operative work with other Agricultural Experiment Stations in the South and with government, industry, and private research agencies.

During World War II statistics was rapidly becoming recognized as a vital branch of science. This recognition came not only because of the developments being made in the theory and techniques of statistics, but also because research workers in many areas were recognizing the scope, usefulness, and efficiency of statistical methods for scientific research. In North Carolina both the Department of Experimental Statistics and the Institute of Statistics were part of these developments.

As the war drew to a close, there were increased state, Experiment Station, and General Education Board funds. Faculty and support personnel at North Carolina State College were added to the point that they could sustain courses for undergraduate and graduate work, as well as expand research and consulting programs. Around this time Robert J. Monroe, Alva L. Finkner, David Mason, Harold F. Robinson, and Jay Wakeley returned from war service. Also Henry 'Curly' Lucas, William G. Cochran, Francis E. McVay, and Paul Peach were added to the staff, and Jackson A. Rigney became a full-time staff member of the

Department of Experimental Statistics. Research support, secretarial services, and book-

keeping activities were strengthened, and the Statistical Computing Service Laboratory was

expanded to include 12 statistical clerks. Approximately 20 graduate students, most of them World War II veterans, were enrolled either as M.S. or Ph.D. candidates.

Expansion of the Institute of Statistics Gertrude M. Cox made several trips to New York in 1944 and 1945 to negotiate with the General Education Board staff, and she traveled to other places to interview prospective staff and to locate potential funds. Included were visits to Columbia University and to Mountain

176

Lake, New Jersey, to talk with Harold Hotelling. As a result of discussions with Miss Cox he agreed to meet with her and University President Frank Porter Graham to talk about develop- ing a Department of Mathematical Statistics at the Chapel Hill campus of the University of North Carolina as part of the Institute of Statistics. Harold Hotelling wrote on a sheet of hotel stationery what staff he would want, with proposed salaries, to start a Department of Mathematical Statistics.

Miss Cox brought back that sheet of hotel stationery and wrote a proposal which was sub- mitted by way of University channels to the General Education Board. On 7 December 1945, they awarded a second grant, this one for $125,000, to supplement existing funds, thus enabling further expansion of the Institute of Statistics. Miss Cox had secured a promise from Professor Cochran that he would join the Institute staff if Dr Hotelling did. The latter had agreed that he would join the program if money was obtained to cover salaries proposed for five years for five professors. Hotelling and Cochran accepted appointments by phone on 8 December 1945.

On 17 February 1946, there was a public announcement that the Institute of Statistics had been approved on an All-University basis by the Consolidated University officials. The Board of Trustees approved formation of the Department of Mathematical Statistics at Chapel Hill to begin 1 July 1946.

Miss Cox continued as the director of the Institute of Statistics, responsible to the office of the President; the associate directors were William G. Cochran (Raleigh) and Harold Hotelling (Chapel Hill). The graduate program and offerings of the two academic departments were closely coordinated. The Department of Mathematical Statistics at Chapel Hill provided strong offerings in probability theory and mathematical statistics while the Department of Experimental Statistics at Raleigh provided strength in intermediate and advanced methods. Throughout the period 1945 to 1960, Ph.D. majors in statistics on both campuses were required to take course work in both departments in order to fulfill degree requirements. To facilitate this interchange of students, classes were scheduled on Tuesdays and Thursdays in Chapel Hill and on the remaining days in Raleigh. Both departments have continued to expand their staffs and course offerings, and today there is still voluntary interchange of students. (See further descriptions of each department later in this article.)

Another development of the postwar period in North Carolina was that the Biometric Section of the American Statistical Association started the Biometrics Bulletin (in 1945), with six issues per year and with Gertrude M. Cox as editor for the first 11 years. In 1947 the name was changed to Biometrics, with four issues a year. The Biometric Society was founded 6 September 1947, as an international society devoted to the mathematical and statistical aspects of biology; it began at once using this journal as its official publication channel. In September 1950, Biometrics became Biometrics, Journal of the Biometric Society.

Regional leadership has been an interesting part of the Institute of Statistics story. One of the stipulations of the General Education Board grants was that the Institute accept respon- sibilities for stimulating activities in statistics in the South. To fulfill this assignment, the Institute provided the following regional services: (a) work conferences, 15 of which were held, both in Raleigh and in other locations, from 1941 to 1948; (b) summer sessions, five of which were held with regional, national, and international participants, 1941 to 1951; and (c) professional assistance. In fulfilling this third area of responsibility, the staff members of the Institute provided help and advice whenever and wherever they could, making visits to other universities to consult on local and regional projects. George W. Snedecor was employed to spend a quarter (or, sometimes, a semester) at each of four colleges in the South besides one year at North Carolina State College. These statisticians did teaching and consulting work and often were asked to make recommendations regarding the organization and coordination of statistics at these colleges.

177

As the postwar period progressed, the Institute of Statistics continued its leadership role and continued to foster cooperative ventures. Some of the earlier cooperative research was conducted with the United States Bureau of Agricultural Economics, Weather Bureau, Office of Naval Research, Soil Conservation Service, Bureau of Plant Industry, Bureau of Animal Industry, Bureau of Mines, Tennessee Valley Authority, and Agricultural Experiment Stations in the South, in the United States, and internationally.

Research in statistical theory was the major function of a number of the staff members. Others were more heavily engaged in developing and testing sampling techniques, experi- mental designs, and methods of analysis to lead to more efficient procedures for obtaining and interpreting quantitative information, e.g., study of breeding and selection procedures to yield most rapid progress in the genetic improvement of animals and plants. All staff members of the Institute were encouraged to maintain close contact with other leaders in the statistical profession, not only by attending and giving papers at scientific meetings but also by visiting other research centers and universities and by correspondence and publications. There was extensive participation in the work of national and international organizations by serving as officers and editors of journals and by performing referee, council, and committee work.

As early as 1943, the Department of Experimental Statistics and the Agricultural Experiment Station provided for research, teaching, and graduate assistants in the area of quantitative genetics. This was made possible when Ralph E. Comstock joined the faculty. In April 1947 the Rockefeller Foundation made a grant of $59,500 to the Institute to develop further the Quantitative Genetics Program, of which Dr Comstock and Harold F. Robinson were the co-leaders. The program was expanded in the early 1950s by the addition of C. Clark Cockerham, Ken-ichi Kojima, Dale F. Matzinger, and Robert H. Moll, as well as several postdoctoral fellows. Another grant, $170,000 from the Rockefeller Foundation, was received later.

Two key events in the history of statistics in North Carolina became official in 1949. One was that on 1 April Jackson A. Rigney became head of the Department of Experimental Statistics, thereby allowing Gertrude M. Cox to devote full time to the work of directing the Institute of Statistics. On 1 July 1949, a third department was added to the Institute, the Department of Biostatistics in the School of Public Health on the Chapel Hill campus, with Bernard G. Greenberg, one of the first Ph.D. recipients in statistics from North Carolina State College, as chairman. The mission of this department has been to provide training and consulting assistance to health-related programs in the University and to fulfill research needs in this area. (See the departmental sections later in this article.)

It was at the Auburn Conference (7-9 September 1948), the fifteenth and last of the series of work conferences sponsored by the Institute of Statistics and the General Education Board, that Gertrude M. Cox presented a 'Proposed Statistical Plan for the South-eastern States'. In the proposal she emphasized that the growing appreciation of the value of statistics had created a demand for adequately trained personnel. Suggestions were made concerning a multistate regional program for coordination of conferences, training statisticians in summer schools, visiting lecturers, and expansion of statistics at each university in the South.

The Southern Regional Education Board (SREB), dedicated to the improvement of graduate and professional education in the region, was established by interstate compact in June 1949. At an SREB-organized Graduate Work Conference which met at Daytona Beach, 4-7 September 1950, one conferee group was instructed to consider 'unique services' in the region. It was this group that first discussed the potentials of a coordinated, cooperative regional program in the then 'unique' field of statistics; a report of the rapidly developing program of the Institute of Statistics was presented as an example of regional cooperation, a role it has continued to serve.

178

Department of (Experimental) Statistics

The early history of the Department of Experimental Statistics at North Carolina State College was closely interwoven with the development of the Institute of Statistics, as described above. In the postwar period the department burgeoned. By July 1949 the faculty had increased to 14 members. H. Fairfield Smith was added to help strengthen the theory offerings. John Wishart, William J. Youden, and George W. Snedecor were visiting professors. There were supportive positions for resident collaborators and assistant statisticians. Most of the time up until 1960, the visiting professors helped with teaching, research, and consulting. They helped broaden the scope of the permanent staff's knowledge about the extensive use of statistical methods and introduced new techniques, especially some of the ones they were finding useful in their own consulting.

Individual consulting work of the staff members had been heavy from the day the program started, especially in biology and economics, and industrial consulting work began during and right after the war. Some examples are: (a) study of the behavior under stress of textile yarns containing both cotton and rayon; (b) seeking the relation between hydrocarbon melting point and size, weight, and conformation of the molecules; and (c) study of sources of error in the determination of amino acid concentrates in biological assay. There were continually increasing numbers of contributions to the statistics and other science subject area journals.

The Department of Experimental Statistics, Raleigh, with the help of the Department of Mathematical Statistics, Chapel Hill, offered course work toward M.S. and Ph.D. degrees. It also offered the baccalaureate degree, but very few students then chose to major in statistics at the undergraduate level. Courses were designed for those preparing to teach statistics and develop statistical theory and for research scholars in other sciences who wanted a practical working knowledge of statistical theory and models. Much of this development was made possible by the General Education Board's rather large grants in 1944 and 1945 which enabled staff to be secured. Many students came to take graduate work in statistics with the help of graduate assistantships and funds from the Veterans Administration and other federal sources. By 1949 six Ph.D. degrees and ten M.S. degrees had been awarded.

Developments of the 1950s The department showed steady development during the 1950s in terms of faculty, graduate students, and financial support. By 1959, the faculty had grown to nine full professors, three associate professors, and seven assistant professors (not including visiting professors and postdoctoral fellows). During the ten-year period, 1950 through 1959, a total of 46 Ph.D. degrees and 41 M.S. degrees were awarded.

Also during this period, substantial funds were being secured from governmental agencies in addition to those already mentioned, including such defense agencies as Office of Naval Research, Air Force, and Army Ordnance (industrial design). In addition to federal and state funds in support of statistical programs, there were industrial contracts undertaken by the Institute of Statistics. On 12 January 1953, University President Gordon Gray announced that financial grants made to the Institute of Statistics just during the last part of 1952 amounted to $606,000.

A basic tenet in the strategy of developing the Department of Experimental Statistics was that it must provide statistical service to all fields in the University and in the region. This was the cornerstone for continued support and growth, and several activities emerged during the 1950s as a direct result, as noted below.

One development was the formation of the North Carolina State College Computing Center. On 1 April 1956, the Department of Experimental Statistics, with support from the North Carolina Agricultural Experiment Station and the Institute of Statistics, acquired the

179

first electronic digital computer in an educational institution in North Carolina (an IBM-650 computer). Arnold H.E. Grandage was named director of the newly formed Computing Center, which was administered by the department. All University faculty were encouraged to make use of the facilities. (Administration of the Computing Center was transferred to the Dean, School of Physical and Mathematical Sciences on 1 July 1962, and from there to the Provost's Office on 1 July 1967.)

On 26 November 1956, Jackson A. Rigney began what turned out to be a two-and-one-half- year leave of absence as department head to serve as chief of mission of the Agency for International Development North Carolina State College program in Peru. His selection for this assignment was the direct result of having become broadly acquainted with agricultural research in his role as plant science statistician. Robert J. Monroe served as acting head of the department during this period.

On 1 July 1958, Alva L. Finkner developed the Southeastern Fish and Game Statistics Project, a cooperative arrangement among the 14 Southeastern states, to provide statistical consulting and applied research to their fish and wildlife program. Guided by the South- eastern Wildlife Commissioners Association through a five-man steering committee, this project is continuing and is growing in breadth. It has been directed by Don W. Hayne since 1962. It supports two-and-one-half full-time faculty members and several graduate research assistants.

The research in quantitative genetics jointly headed by Ralph Comstock and Harold F. Robinson attracted international attention during the 1950s and resulted in an organizational change. On 1 July 1958, the major contingent of the Quantitative Genetics Program was separated from the Department of Experimental Statistics to form the nucleus of the new Department of Genetics, in the School of Agriculture and Life Sciences. Harold F. Robinson, Professor of Statistics, became the head of the new department, and Professor of Genetics. C. Clark Cockerham, Professor of Statistics and Genetics, remained in the Department of Experimental Statistics to continue his research and to provide consultative assistance in statistical genetics and plant and animal breeding to the other departments.

Developments since 1960

Developments proceeded at the same fast pace in the 1960s. In the post-Sputnik era, there was considerable agitation to give the basic sciences more visibility and autonomy on the North Carolina State College campus. One such move was the formation 1 July 1960, of a School of Physical Sciences and Applied Mathematics consisting initially of four departments: Chemistry, Applied Mathematics, Physics, and Experimental Statistics. The departments of Chemistry and Experimental Statistics were taken from the School of Agriculture, and the departments of Applied Mathematics and Physics were moved from the School of Engineering. Although administration of statistics teaching and associated academic activities was shifted to the new school, a Department of Experimental Statistics remained in the North Carolina Agricultural Experiment Station because of research projects directly associated with that area and to provide the service functions of statistical advising, consulting, and computing services.

With the transfer of academic activities to the new school, the undergraduate curriculum in statistics was revised to be more attractive to the student oriented to mathematical and physical sciences. Since the early 1960s, 20 to 35 undergraduates per year have majored in statistics.

Throughout these changes, the Institute of Statistics continued to support the major portion of the theoretical and applied statistical and biomathematical research work with state, grant, and contract funds.

Some reorientation and development in the 1960s culminated in 1965 with a change in the name of the institution to North Carolina State University at Raleigh. In 1970 the name of the school became School of Physical and Mathematical Sciences, and concurrently, the name

180

of the Department of Experimental Statistics was changed to the Department of Statistics. This change in name was in recognition of the department's increase in theory-oriented faculty, thereby providing a full program in statistics.

The use of statistical concepts and designs in a variety of biological research programs led inevitably to a strong interest in quantitative modeling of biological phenomena. The Biomathematics Program, with H.L. Lucas as director, was formally recognized with the awarding of a National Institutes of Health Training Grant in Biomathematics on 1 January 1961; actually Dr Lucas and several of his colleagues had been involved in the mathematical modeling of biological systems for several years. H.R. van der Vaart was employed in 1961. In the summer of that year the Biomathematics Program sponsored a two-week international conference in biomathematics at Cullowhee, North Carolina. This highly successful conference not only stimulated more interest in biomathematics in North Carolina, but throughout the nation and the world.

On 1 May 1966, the Biomathematics Program received further stimulus with the award of a $1.125 million National Science Foundation Science Development Grant over a three-year period. Emphasis in the Biomathematics Program continues to be on graduate training and research, and support has been provided for several faculty, postdoctoral fellows, and an average of 12 to 15 predoctoral fellows per year.

On 1 March 1962, Jackson A. Rigney accepted a temporary foreign assignment and David D. Mason was named acting head of the Department of Experimental Statistics and acting director of the Raleigh section of the Institute of Statistics. Later in the year Professor Rigney elected to continue in international activities. Professor Mason, effective 1 July 1963, formally assumed both posts.

Another event of the 1960s was formation of the Joint Econometrics Program. There had

long been informal cooperation between the departments of (Agricultural) Economics and

(Experimental) Statistics. This was facilitated by the fact that they were housed in the same building. T. Dudley Wallace was jointly employed by the two departments on 1 September 1959. The Joint Econometrics Program, which grew from Dr Wallace's activities, was formally initiated in 1964.

The program continues to be active in 1976, with two joint faculty appointments and several other members from both the Department of Economics and Business and the Department of Statistics heavily involved. Professor Wallace resigned in 1974 and was replaced by Thomas Johnson. A.R. Gallant, appointed in 1971, replaced Charles H. Little.

The Quantitative Genetics Program, cooperative between the Departments of Statistics and Genetics, was given a big stimulus in 1963 with the award of a National Institutes of Health Program Project Grant in the amount of $2.2 million for seven years. The already well-known program has since achieved worldwide recognition. The grant is currently in its thirteenth year and supports work in both the Department of Genetics and the Department of Statistics.

Leadership in statistical computing developed during the 1960s. Since the formation of the department in 1941, emphasis has been placed on providing service computing for the research staff, especially the Agricultural Experiment Station. First, the major portion of the computing was done on desk calculators, with as many as 12 clerks employed at one time. As mentioned earlier, the Department of Experimental Statistics acquired an electronic digital computer (an IBM-650) in April 1956, the first in an education institution in North Carolina and one of the first in the nation.

The acquisition required pioneering in writing and documenting computer programs for summarizing and analyzing research data. Considerable assistance was afforded this effort by a National Institutes of Health computer grant initiated in June 1962 and carried until 1 October 1973, at an annual amount ranging from $120,000 to $140,000. Through assistance from this grant, the IBM-650 computer was replaced by the larger, more versatile IBM-1410 in 1963.

181

The most radical change in computer equipment that affected statistical computing came in 1965, with the introduction of the IBM-360 series. This series required the rewriting of all programs, especially those that utilized assembly language. This put many universities in a difficult situation. In the Southern region, through the organization of the University Statisticians of the Southern Experiment Stations, a committee was formed with Francis J. Verlinden of North Carolina State University as chairman to coordinate efforts in the development of statistical software for data analysis. Since the Department of Experimental Statistics had more programming personnel, and the support of the computer grant, it accepted the major portion of the task of re-programming.

The concept of the Statistical Analysis System (SAS) was initiated in 1966 with the re- employment of A.J. Barr; he had just spent two years working in Washington, D.C., for IBM Federal Systems. Shortly after Mr Barr started the SAS development, J.H. Goodnight began working part time as a student, and later full time, on the statistical routines. Mr Barr and Dr Goodnight have consequently emerged as co-leaders of the SAS Project.

Upon termination of the computer grant support, the SAS project was accepted (effective 1 July 1973) as a regional project by the Southern Experiment Stations. This support plus that from other users has enabled the Department of Statistics at Raleigh to maintain a strong position in the area of statistical computing. On 1 July 1976, the core personnel of the highly successful SAS Project left the University to form a private software corporation, the SAS Institute, Incorporated.

Summary Highlights of Consulting and Research The Department of Statistics at North Carolina State University has been heavily involved in statistical consulting and advisory activities since its formation in 1941. The first consulting was with research personnel of the North Carolina Agricultural Experiment Station, and this continues to be the heaviest consulting responsibility of the department, involving plant and animal sciences, economics, sociology, biological and agricultural engineering, the biological sciences, and food science. Complete service is provided, from planning and design of experi- ments and surveys, to collection and analysis of data, to review of manuscripts.

As the faculty and interests of the department grew, so did the range of statistical con- sulting services provided to the University and state agencies. In addition to the Experiment Station services, statistical consulting is provided for the engineering and physical sciences, education, behavioral sciences, and various other disciplines. An example of one large responsibility is the previously mentioned 18-year-old Southeastern Cooperative Fish and Game Statistics Project, through which statistical consulting, computing, and applied research services are provided to the Southeastern states in estimating game kills, fish caught, fish and game population estimates, and other services necessary for the management and regulation of wildlife resources.

Faculty members heavily involved in the department's consulting services often become involved in the research activities as well - e.g., developing and improving methodology and adapting existing methods of data analysis to particular situations. They also share in the publication of results. The department's lively research activity is reflected in the publication of 25 to 35 refereed journal articles per year plus numerous reports, presentations at professional meetings, and special conferences.

The research can be grouped in the more or less traditional categories. In the area of estimation, testing, and decision procedures, there has been considerable activity in estimation of variance components, regression analysis including use of prior information, minimum bias estimation of polynomial response relationships, multivariate theory and analysis, non- parametric methods (including nonparametric discrimination), maximum likelihood estimation for true-type scales, and studies in goodness-of-fit using transformation methods.

182

In the area of experimental design, there has been a continuing interest in response surface design since George E.P. Box visited the department in 1952-53, and generally in the development and improvement of designs and accompanying methodologies for use in agri- cultural, biological, and industrial experimentations. Recently there has been investigation of response surface fitting (or trend analysis) to account for variation in experimental material in addition to or in lieu of blocking. Investigation of the choice of mathematical models for soil fertility data and consequent optimal experimental design has been an active area.

Sampling theory has been of continuing interest throughout the entire period. The various aspects of multistage sampling, including sampling on successive occasions and replicated samples of unequal size, have been studied. The basic problem of model construction in measurement and sampling of social networks is still under investigation. On the more applied side, there has been continuous activity for at least 20 years in development and refinement of theory and methodology in the design and analysis of sample survey and sampling methods in the agricultural, biological, and industrial areas.

Research and training programs in biomathematics and in quantitative genetics have already been mentioned. In both areas the basic thrust has been toward mathematical modeling of biological systems. The Quantitative Genetics Program has been and is joint with the Depart- ment of Genetics, which has been primarily concerned with experimental research whereas the

Department of Statistics has been oriented toward theory development. An exception to this has been in the area of numerical taxonomy, which has involved both theory and experi- mentation.

All in all, the faculty of the Department of Statistics at North Carolina State University has covered a wide range in research interests, including rather theoretical contributions and

highly applied methodology.

Department of (Mathematical) Statistics

As described earlier, the formation of the Department of Mathematical Statistics at the

University of North Carolina at Chapel Hill followed the expansion of the Institute of Statistics to All-University status in 1946. With the development of this second department, training and research in statistics in North Carolina proceeded at a more rapid pace.

Developments up to the Mid-Sixties In 1945-46 while negotiations with Harold Hotelling and the securing of additional financial support was progressing, the Institute of Statistics hired Jacob Wolfowitz to give basic graduate probability and statistical inference lecture courses, with considerable attention to philosophical implications. After one year of pioneer teaching, on both the Chapel Hill and Raleigh campuses, Wolfowitz decided to return to Columbia University in order to work with Abraham Wald.

On 1 July 1946, the Department of Mathematical Statistics in the Graduate School was officially started under the chairmanship of Harold Hotelling, who assembled a faculty con- sisting of William G. Madow, P.L. Hsu, and Herbert E. Robbins. Maurice S. Bartlett was a visiting professor and Edward Paulson an instructor.

Dr Hsu, Dr Madow, and Dr Paulson left in the next two years, but the faculty was strengthened in the fields of experimental design and multivariate analysis by the addition of two outstanding Indian statisticians, R.C. Bose and S.N. Roy, and in probability theory and statistical inference by Wassily Hoeffding, a young worker of great promise, which has since been abundantly fulfilled, and who is still with the department. Walter L. Smith, to whom the same description fully applies, came from England to join the department in 1954, following the departure of Dr Robbins.

183

The department benefited greatly during the formative years from the able guidance of Harold Hotelling. It was from his mature judgment that the main lines of development were laid down. One of his many capabilities was in the area of multivariate analysis. In co- operation with Gertrude M. Cox, Harold Hotelling wrote a proposal for a large research project in the areas of multivariate analysis, probability, and other mathematical statistics areas. A large grant was received from the Department of the Navy, Office of Naval Research, as early as 1947, and substantial funds were received from this source during the rest of Hotelling's active research years. These funds provided supplementary help in securing other key staff members, such as Wassily Hoeffding and S.N. Roy. Also, funds were provided for visiting professors and numerous graduate students. During the period 1948-55 there were at least 40 reports published in the Institute of Statistics series and in statistical journals, especially the Annals of Mathematical Statistics.

Very soon after securing basic research funds from the Office of Naval Research, a contract was secured in 1951 from the Air Force, which substantially helped develop work in sequential design with R.C. Bose as project director. This contract helped secure new staff members, visiting professors, and graduate assistants. From this Air Force contract, the Institute of Statistics received over $168,000 between 1 November 1951, and 30 June 1957.

Wassily Hoeffding was director of an Air Force project on nonparametric statistics which

produced 11 articles submitted for publication to Annals of Mathematical Statistics, Journal of the Royal Statistical Society, Biometrika, and Biometrics in one year, October 1953 to October 1954.

These are only three of the many contracts the Institute of Statistics had with the Office of Naval Research, Air Force, and Army Ordnance Corps. The capable staff secured by Harold

Hotelling for these projects, the reports, publications, and graduates made Chapel Hill a world center in these fields.

Harold Hotelling was succeeded as chairman in 1952 by George E. Nicholson, Jr., who continued existing policies and developed new ones. His interest in applications of statistics in public policy and operations research type of inquiries was reflected in his multi- farious consulting activities; it introduced a component into the work of the department that survives to the present day. Also in 1952, the official name became the Department of Statistics.

The research output of the department in the 1950s was quite remarkable, including in 1958-59 the well-known work of R.C. Bose with S.S. Shrikhande, a visiting professor, on the resolution of Euler's conjecture on the existence of Graeco-Latin squares, a problem of nearly 200 years' standing. The department's tenured staff remained small in this period, but there were several visitors to the department, which fact attests to the importance of the research work then in progress.

Under the leadership of Walter L. Smith, rapid progress was being made at that time in

queueing theory; eventually, in 1964, there was a successful conference on congestion theory sponsored by the department in cooperation with the Office of Naval Research.

Beyond its main endeavor, research, the department had considerable influence on the

development of statistical work on the Chapel Hill campus. Writing in the annual report of the department in 1968, Dr Nicholson, who was an active agent in much of this, remarked: 'The philosophy of the department was to cooperate and not compete with the development of statistics in all its aspects and to strengthen existing programs and to aid in establishing new ones. Consequently the department acted as the executive agent for the Institute of Statistics in its efforts to assist in establishing the Department of Biostatistics in 1949; cooperated with the Institute for Research in Social Science to establish the Social Science Statistical Laboratory in 1956; with the Graduate School to establish the Department of Information Science in 1964; with the Medical School to establish a biomathematics program in the School of

184

Medicine in 1965; and with the University to organize and participate in the Triangle Universities Computation Center also in 1965.'

By 1965 some staff changes had occurred. A very sad one was the untimely death, in 1964, of S.N. Roy. This was a blow to the multivariate work of the department, which was further weakened by the retirement in 1966 of Professor Hotelling. However, advances were being made by newly recruited faculty. Thus, William J. Hall joined the permanent faculty in 1959, Norman L. Johnson in 1962, I.M. Chakravarti in 1964, and Malcolm R. Leadbetter in 1965.

Dr Hall further strengthened the department in probability theory and Dr Chakravarti contributed to the further development of the new fields of coding theory and information theory to which R.C. Bose had turned his attention in the late 1950s. Bose and Chakravarti also continued the work of the department in experimental design, in particular with respect to the use of combinatorial mathematics and finite geometries in constructing such designs. Dr Leadbetter's initiatives in the field of stochastic processes and statistical inference were notable. His book co-authored with Harald Cramer should be mentioned here (Cramer and Leadbetter, 1967).

Developments since the Mid-Sixties In 1966 the Department of Statistics became a regular department of the College of Arts and Sciences, University of North Carolina at Chapel Hill.

The distinguished work in teaching and research continued. A summer conference on combinatorial mathematics and applications, organized by the department in cooperation with the Institute of Statistics of the University of Paris and sponsored by the North Atlantic Treaty Organization (NATO), was held at Royan, France, in 1965, and a symposium with the same title, sponsored by the United States Air Force Office of Scientific Research, was held at Chapel Hill in 1967. Work in this field culminated in a Combinatorial Mathematics Year (actually 16 months, from 1 February 1969, to 31 May 1970) sponsored by the National Science Foundation (NSF) and the United States Air Force Office of Scientific Research. There was a symposium in June 1969 and a second conference on combinatorial mathematics and

applications in May 1970 in Chapel Hill. During the Combinatorial Year there were about 20 special visitors in the department.

A feature of the later 1960s was the development of a strong stochastic processes program, led by Walter L. Smith and Malcolm R. Leadbetter. They succeeded in arousing continuing interest in this program among their colleagues, and it is now an important facet of the work of the department. An NSF regional symposium on multiple time series in January 1973

(principal speaker E. Parzen) was one outcome of this activity. A considerable expansion in the work of the department, and a rapid increase in number of

faculty occurred in the period 1968-70. This was stimulated by an NSF Science Development (Center of Excellence) grant which operated from 1968-73, with extension to 1975. The Department of Statistics at Chapel Hill, one of six in the United States, benefited considerably from the grant, although some of the benefits proved unexpectedly transitory because the University was unable to honor its commitment to continue the funding of some of the new positions created by the grant.

In 1968, with the appointment of Charles R. Baker, courses in statistical communication theory were introduced. This program was strengthened by the appointment of Stamatis Cambanis in 1969 and has developed steadily. Floyd J. Gould joined the faculty in 1968; he had special interests in operations research, a field in which, as already noted, George E. Nicholson was also strongly interested. There has been a separate Curriculum in Operations Research and Systems Analysis since 1973, but the Department of Statistics continues to offer an operations research option to its own graduate students.

Other faculty recruited at this time include Gordon D. Simons, the present chairman;

185

Douglas G. Kelly, joint with the Department of Mathematics; K.J.C. Smith; and Edward J. Wegman; all are still in the department in 1976. William S. Cleveland, Thomas Allan Dowling, and Woolcott W. Smith also joined the faculty but left within a few years.

In 1968-71, in cooperation with the Department of Biostatistics and the Psychometric Laboratory, the department participated in a program sponsored by the National Science Foundation to study the effectiveness of using computers in undergraduate statistics instruction. At about this time, 1969-70, courses intended specifically for undergraduates were introduced. Before then the department had offered no undergraduate courses, though a few undergraduates did attend some courses in the department, and it was possible to take a statistics major in the bachelor of science degree program.

The development of undergraduate courses in the department has been striking. In 1969-70 there were about 15 undergraduate students enrolled for courses in the fall semester and 40 in the spring. The corresponding figures for 1975-76 were 518 and 540. This very considerable expansion has resulted, since 1972, in increasing opportunities for the graduate students to obtain teaching experience. Initiation and continuation of this expansion owes much to K.J.C. Smith, who devoted considerable effort to formulating and publicizing the courses. The further utilization of computer resources in undergraduate instruction has received continued attention, especially by Edward J. Wegman.

In 1971, George E. Nicholson, Jr., was succeeded as chairman of the department by Norman L. Johnson, who in turn was succeeded by Gordon D. Simons in 1976. The department suffered a major loss with the retirement of R.C. Bose in 1971. His outstanding contributions to experimental design, coding theory, and information theory provided inspiration for much research activity in the department. Unfortunately, the department has not been able to replace him. I.M. Chakravarti and Douglas G. Kelly have continued this work in a most able manner, but a full-time co-worker in this field is needed.

The department has suffered also from the lack of an experienced multivariate analyst since the death of S.N. Roy in 1964. It is only due to the generosity of Pranab K. Sen and the Department of Biostatistics in making his time available both for lecturing and student guidance during the last ten years that a viable multivariate analysis program has been continued to the present.

The latest addition to the faculty is R.J. Carroll, who joined the department in 1974. His field of specialization is selection procedures, and he has contributed effectively to the teaching and consulting activities of the department.

Since the earliest days, the faculty members of the department have played active roles in the affairs of national and international statistical societies by way of being office holders, and they have served on committees and on editorial staffs of journals. These activities are so numerous and widely recognized that no attempt is made here to set them out in detail. Four present or previous faculty members have been elected to the National Academy of Sciences.

Students in the Department of Statistics

The varied interests of the faculty make it possible to offer quite a wide choice of specializations to graduate students in the Department of Statistics at Chapel Hill. At present they have a choice of six major fields: inference, design, multivariate analysis, probability, communication theory, and operations research. Whichever field is chosen, students are required to take several broadening courses in other fields and in other departments.

Since its inception, the courses offered by the department have had a strong mathematical and theoretical flavor. Still, the department has always operated on the principle that appli- cation and theory can not be separated and, indeed, that each stimulates the other. This was partially recognized as early as 1952 when the name was changed from Department of Mathematical Statistics to Department of Statistics. Indeed, members of the faculty have

186

worked on applied statistical problems throughout the lifetime of the department, but this was not reflected in courses available to students until recently. In the last two years an experimental course in problems of applied statistics has been offered. It is not planned that this should be a major activity of the department; it is intended to assist in the general statistical education of graduate students in the department.

The department has been very fortunate in the quality of its graduate students. Not only has each year's entry class been among the highest in the University in terms of conventional test scores and grade-point averages, but this promise has on the whole been borne out by subsequent performance.

Number of graduate students have remained fairly stable over the last ten years. In fall 1976, there were 40 graduate students enrolled, a figure somewhat higher (by 3-5 students) than those over the previous five years.

The first Ph.D.s were awarded in 1948. Up till 1 September 1976, 62 M.S. degrees and 125 Ph.D.s had been obtained in the department. After the initial period of growth, the number of Ph.D.s awarded has remained more or less stable at about four to six per year since 1954.

Department of Biostatistics

Still another vital element in the distinguished history of statistics in North Carolina is the contribution of the Department of Biostatistics which developed at the University of North Carolina at Chapel Hill (UNC-CH). It has been said: 'It is generally recognized ... that today the biostatistics department at the University of North Carolina has one of the most advanced programs in the nation' (Sheps, 1976).

Early Developments The Department of Biostatistics was established in the School of Public Health on 1 July 1949, with the active collaboration and financial support of the Institute of Statistics. Early on, Edward G. McGavran, dean of the school agreed that one of the requirements for the key appointment of chairman of the new department was a person thoroughly familiar with mathematical statistics who could develop joint programs both with Harold Hotelling in the Department of Mathematical Statistics and Gertrude M. Cox in the Department of Experi- mental Statistics in Raleigh. Such a leader was found in the person of Bernard G. Greenberg who himself now serves as dean of the school.

The primary mission of the new department was to teach one or two service courses and to provide consultation in statistics to meet the needs of students and faculty in public health and the other health disciplines in the UNC-CH Division of Health Affairs. It soon became clear that, without research and without training biostatisticians, the service role in teaching and consultation would become sterile and obsolete. Accordingly, a training program leading to the M.P.H. degree was introduced in 1950 and the first student graduated in 1951.

From an initial core course in elementary public health statistics, the main teaching function expanded to include specialized courses in biological assay, statistical methods in epidemiology, and the history and uses of vital statistics and demography in the administration of health programs.

In 1951, responsibility was assumed for teaching statistical methods to medical and dental students. This increased load brought a transfer of funds from those two units, which enabled the department to employ C. Clark Cockerham the following year. Cockerham, a new Ph.D. from Iowa State in genetics and statistics, remained for one year and concentrated upon teaching and consulting with medical students and faculty. He transferred in 1953 to the statistics and genetics program in Raleigh and has continued to be a bulwark of strength for

187

that effort. He was replaced by Harry Smith who was a graduate student in the Raleigh program at the time.

When Smith became an instructor in 1953, the year was marked by two key developments in the Department of Biostatistics. The first was the formal introduction of a joint educational effort between Biostatistics and the other two statistical departments in the Institute. Instead of granting its own Ph.D. degree, the plan was to enroll the doctoral students from Biostatistics in one of the other two statistics departments for fulfilling the formal degree requirements. Courses in public health and/or basic preclinical medical subjects were individually prescribed for each student, and direction of the doctoral dissertation could be undertaken by a faculty member from any one of the three departments. Thus, Dr Greenberg served as chairman of the doctoral committee for Harry Smith who received the first such degree in 1954 from Raleigh.

This arrangement has continued to the present time, although it was considerably reduced in 1965 when Biostatistics applied for and received permission to enroll students for its own Ph.D. degree through the Graduate School. By that time, the number of faculty in Biostatistics had grown to a size where the department could be independent in supervising doctoral students.

The second key development in 1953 was the receipt of a training grant from the National Heart Institute of the National Institutes of Health, United States Public Health Service, to train biostatisticians to serve as consultants in health research. This grant was one of the original four made to departments of biostatistics; this marked the first time that the National Institutes of Health provided funds for any kind of training program. The funds came from the research allocation to the National Heart Institute, and the necessary Congressional committees authorized the use of research funds for training because biostatisticians would enhance the quality of research in the cardiovascular field. After the Russian Sputnik on 4 October 1957, training programs became common in many biomedical disciplines, and the National Institute of General Medical Sciences was established to administer these training grants. The grant in biostatistics to the University of North Carolina School of Public Health is still active after more than 20 years.

This training grant provided funds for student stipends, postdoctoral fellowships, faculty, and one visiting faculty member each year. As a result, the Department of Biostatistics was able to attract as visiting professors for one or more years such outstanding statisticians as David R. Cox, J.O. Irwin, David B. Duncan, David J. Newell, Ahmed E. Sarhan, Herbert A. David, Robert C. Elston, Manindra N. Ghosh, K. Ruben Gabriel, Pranab K. Sen, and others, some of whom remained as regular faculty members.

In addition to the doctoral program in conjunction with the other two departments in the Institute of Statistics, the Department of Biostatistics offered the Ph.D. degree in Public Health, with specialization in Biostatistics, for those who were anxious to emphasize the applications of statistical methods to solve public health problems. The first graduate of this program was Ahmed E. Sarhan in 1955. This particular program was discontinued in the late 1960s, and in 1970 the Dr.P.H. degree was substituted for it. Sarhan remained for an additional year after graduation performing research and working in collaboration with Greenberg on the monograph, Contributions to Order Statistics, which was published a few years later by John Wiley and Sons (Sarhan and Greenberg, 1962).

Two major contributions to the field of statistics made by the Department of Biostatistics prior to the 1960s should be mentioned here. First was the concept of field training, a kind of practicum lasting approximately ten weeks in which the master's degree student is assigned to a field counselor in an operating health agency. The purpose was, and still is, to familiarize students with real problems and to provide individual opportunities to learn more about public health and to apply academic knowledge of statistical methods in solving a health

188

problem. The field training was conceived as an integral part of the curriculum, involving academic credit and supervised by regular faculty as well as the field counselor.

The initial collaborative arrangement was developed with the Tennessee State Department of Health in Nashville under the field direction of Ruth Puffer and Ann Dillon. Later, it was expanded to include local and federal health units as well as other state departments of health, organizations for prepaid medical care, and health research agencies. The role of field training is explained in more detail in a publication by Greenberg (1964) in The American Statistician in February 1964.

The second contribution of the department in this period involved the pioneering efforts of the faculty in developing the principles and practices underlying the cooperative, multicenter, clinical trial. In 1955, the Department of Biostatistics undertook a contract with the Cancer Chemotherapy Program of the National Cancer Institute to serve as the statistical co- ordinating center for the Southeastern Cooperative Cancer Chemotherapy Study Group, a consortium of approximately ten medical schools conducting joint research with chemo- therapeutic agents believed to be of value in treating leukemia and lymphoma. James E. Grizzle, now chairman of the department, joined it a year later as a research assistant under this program and helped Bernard G. Greenberg to establish the philosophical principles for organizing and statistically servicing collaborative clinical trials. Later, the program was extended through a contract with the Veterans Administration to provide statistical services to Veterans Administration hospitals studying the relative value of four operations in the treatment of gastric ulcer. In 1959, Greenberg (1959) published in The American Statistician the first available article describing the design and conduct of cooperative field and clinical trials.

In 1958, there were two important developments that influenced the subsequent develop- ment of the department. Roy R. Kuebler, a new Ph.D. recipient in mathematical statistics and biostatistics, who worked under the supervision of R.C. Bose, remained with the department on a permanent basis. He added a new dimension to the teaching efforts of the Department of Biostatistics and was a source of strength in building the foundation for the expanded degree programs described in the following section.

The other key milestone of 1958 started in 1955 with the return of H. Bradley Wells for a doctoral program in public health. In 1958, he combined this period of study with a work effort involving the North Carolina State Board of Health. Under a contract between the department and the state health department, Wells spent one day a week in Raleigh consulting with program directors on statistical problems. That contract with the state health department for statistical services from a faculty member has continued unabated; Wells was followed by James R. Abernathy, Craig D. Turnbull, and Roger Grimson, who is currently performing this function.

Developments after 1960 The period after 1960 was marked by a rapid expansion in training and research efforts Nothing was more impressive, however, than the successful execution of two special summer sessions in 1963 and 1964. Under a joint arrangement with the Biometry Training Program directors and the National Institute of General Medical Sciences, the Department of Bio- statistics sponsored each summer a special session with outstanding visiting professors teaching 10 to 12 courses in biostatistics, ranging from elementary courses to the most advanced possible. Well over 100 students enrolled during each summer session to take advantage of this unique opportunity. The faculty who were assembled to teach at these sessions are listed in Exhibit XIII.

Research grants and contracts continued to grow, providing the necessary funds for expansion of the department. Some grants were for training purposes; others were for basic

189

statistical research, such as order statistics with the Army Research Office; others were for statistical consulting services in cooperative clinical trials. The number of graduate students specializing in biostatistics grew exponentially and the doctoral program was as heavily subscribed as the master's degree programs. Some key dates in inaugurating new educational programs, with accompanying training grants, were as follows:

1965 - Ph.D. in biostatistics authorized by the Graduate School. 1965 - M.S.P.H. and Ph.D. programs in statistical aspects of demography and population

studies. 1965 - Ph.D. program in statistical genetics. 1967 - M.S.P.H. program in biostatistics with special emphasis on problems of mental

health, stressing data analysis and evaluation of programs. 1970 - Ph.D. program, in conjunction with Department of Environmental Sciences and

Engineering, on statistical problems in environmental health. 1970 - Master's, Ph.D., and Dr.P.H. programs to study the problems of health services

research and evaluation of service programs. 1973 - Master's and doctoral programs with concentration in public health data manage-

ment and computers. 1976 - Undergraduate degree in public health (B.S.P.H.) with specialization in biostatistics

as one of the options. In all of the above areas, the Department of Biostatistics has active research work in progress

and is heavily engaged in consultation with health research workers in diverse disciplines. The department has a long record of providing statistical and data-management services to health- related investigations dating back to studies of growth and nutrition and the early cancer chemotherapy clinical trials. Development of new statistical methods has been an integral part of the departmental research program. Major problems have included the design of co- operative clinical studies, fetal and neonatal mortality studies, blood-bank inventory policy, human genetics, order statistics, population and fertility models, multivariate analysis, non- parametric methods, survivorship analysis, and use of randomized response in survey research.

It would be too lengthy a compilation to attempt to identify the above contributions with the individual faculty members who joined the department in the 1960s and thereafter. Never- theless, some of the outstanding faculty who made significant contributions were Dana E.A. Quade, Robert C. Elston, Herbert A. David, Forrest E. Linder, Mindel C. Sheps, Regina E landt-Johnson, Peter A. Lachenbruch, Gary G. Koch, and many others.

Major current efforts in the mid-1970s are:

1. The Central Patient Registry and Coordinating Center for the Lipid Research Clinics Program. This program, which includes two major studies, is designed to measure blood lipids in selected population groups and to test the validity of the theory that lowering cholesterol levels may have a significant effect on heart disease. The data are being gathered in 16 participating clinics in the United States, Canada, Israel, and the Soviet Union.

2. The Study of the Efficacy of Nosocomial Infection Control. This research is designed to extract information on hospital-acquired infections from 500,000 patient charts selected randomly from 350 hospitals.

3. The Study of the Health of Workers in the Rubber Industry. The purpose of this study is to identify work-related illnesses and environmental hazards and ultimately to develop a system of surveillance of the health status of workers and the environment in the rubber industry. This collaborative effort also involves the Department of Epidemiology and the Department of Environmental Sciences and Engineering in the School of Public Health. Both the Lipid Research Program and the Occupational Health Studies Program have associated areas of graduate and postgraduate training.

46/2-E

190

4. The Improvement of Statistical Methods to Measure Population Change. Research in demographic methods has been centered around a cooperative international program called POPLAB (Laboratories for Population Statistics), which is designed to increase institutional capacity in various areas of the world for the collection and analysis of demographic data.

The current department budget is in excess of $3 million, and there are about 75 students, equally divided in master's and doctoral programs. The number of regular full-time faculty is in excess of 30 and has reached a plateau during the last few years. All of this growth to become the largest health statistics and biostatistical training unit would not have been possible without the early active support and collaboration of the Institute of Statistics and the encouragement of Gertrude M. Cox and School of Public Health Dean Edward G. McGavran.

At one period in the late 1960s, the department was simultaneously the host for the editors of Technometrics and Biometrics and associate editors for International Statistical Review, Journal of American Statistical Association, and Journal of Chronic Diseases. Graduates and faculty from the department have been journal editors and contributors; they have established independent reputations for themselves as academicians, department chairmen, administrators, and research scientists.

Such achievements have helped to establish a Department of Biostatistics at the University of North Carolina which will continue to undertake a vital role in developing new statistical methodology, as well as in training and helping to apply good statistical methods by means of consultation with research workers and service agencies. The department has taken its rightful national and international place with the two other major departments in the Institute of Statistics.

Other Ventures

Throughout the history of the development of statistics in North Carolina, there has been a prevalent spirit of cooperation. Much of this tradition has its roots in the facilitating role of the Institute of Statistics. A few more developments in the state should be mentioned.

One is the highly respected L.L. Thurstone Psychometric Laboratory which originated in 1952 in affiliation with the Department of Psychology at the University of North Carolina at Chapel Hill. The laboratory was started by L.L. Thurstone, who was influenced in his decision to come to North Carolina by the existence of the Institute of Statistics.

There was a cooperative relationship with the Institute of Statistics from the beginning. The Institute provided all the salary of T.E. Jeffrey as an assistant for L.L. Thurstone for at least three years and secured a grant from the United States Army Ordnance for cooperative psychological research. This grant provided for statistical staff consulting time, a graduate assistant, and computer machine time for analytical help.

In reminiscing about a later period, Lyle V. Jones, who became director of the Laboratory in 1957 (and who is now Dean of the Graduate School), comments on 'a number of joint activities' involving the Institute and one or more of the academic units.

One of these was the pooling of funds in 1957-58 from the Psychometric Laboratory and the departments of Biostatistics and Statistics to purchase a small computer, the Royal-McBee LGP-30, that was shared by all three units. This was the first high-speed electronic computer on the campus in Chapel Hill. Also a joint effort was a five-year research project on multi- variate analysis, with the principal investigator located in Biostatistics and faculty salaries of selected members in all three academic units coming from funds in the grant.

Another development began when the staff of the sampling section, headed by Alva L. Finkner, Department of Experimental Statistics, Raleigh, began cooperating with Daniel O. Price of the Institute of Research in Social Science, Chapel Hill. The Institute of Statistics

191

secured an Air Force grant for human resource research, and this grant provided part of the salary of Daniel O. Price and the salaries for two graduate assistants. It was decided in 1954 that a Survey Operations Unit was needed to augment the survey research capabilities on both campuses. It was located in Venable Hall, Chapel Hill. John Monroe came from a similar facility at Iowa State University to head the unit. Daniel O. Price and J. Stacy Adams, a doctoral candidate in social psychology (now Reynolds Professor, School of Business Administration) held part-time positions in the first years of operation of the unit. The unit became the Social Science Statistical Laboratory in 1956.

Close relations with all interested departments were assured by a steering committee con- sisting of representatives from the Institute of Research in Social Science and from political science, journalism, business administration, biostatistics, and both departments of statistics. This committee provided guidance in the formulation of policy and practices as well as sources for sampling and survey projects.

In cooperation with the sampling section of the Department of Experimental Statistics, the unit aggregated and developed master sample materials for the state of North Carolina. Most of this work involved counting culture (houses) on county highway maps and creating 'count units' (road-bound areas) to make a frame for area probability sampling of the state, counties, or groups of counties.

Early projects of the unit were trade area studies of Winston-Salem and Norfolk, two data collection projects for TVA studies, a sociological study in Appalachia, and a national sample of veterinarians. Three major undertakings were a study of smoking habits of all American Tobacco Company employees, a highly concentrated health survey in Nashville, Tennessee, and the beginning of an evaluation of the Census of Agriculture in 1959.

During the period 1954-59, two books Interviewing Techniques by J. Stacy Adams (1958) and Handbook of Area Sampling by John Monroe and Alva L. Finkner, and a number of research reports in the Institute of Statistics (1976) series were published. John Monroe organized and taught for several years a course in survey methodology which was designed primarily for a special MPH curriculum in biostatistics.

In March 1959, the Social Science Statistical Laboratory was transferred to the emerging Research Triangle Institute to provide a nucleus of operations personnel for the Social and Statistical Sciences groups. With a combined interdisciplinary staff of more than 250, they have a capability that is distinctive among national survey research organizations.

The Research Triangle Institute is a significant example of the close working ties among the academic centers in North Carolina - North Carolina State University at Raleigh, the University of North Carolina at Chapel Hill, and Duke University at Durham. Under the leadership of the Honorable Luther H. Hodges (North Carolina's industry-minded governor), a Research Triangle Commission was appointed in 1955.

In January 1957, this Commission appointed an Executive Committee to study the problems involved in establishing a Research Institute. This committee invited Gertrude M. Cox to submit to them a proposal regarding the possibilities of establishing a statistical unit in a Research Institute. A committee of seven senior statisticians, with Gertrude M. Cox as chair- man and Alva L. Finkner as secretary, submitted their report 16 June 1958, to the Executive Committee of the Governor's Research Triangle Commission. On 1 July 1959, Gertrude M. Cox (on one-half time) began to organize and staff a Statistics Research Division of the Research Triangle Institute.

The Social and Statistical Sciences groups, with ten centers, have on their staffs a large number of statisticians and social scientists, and have conducted many cooperative research projects with the three academic departments of statistics. Some senior statisticians at Research Triangle Institute hold adjunct professorships and direct the training and provide employment for graduate students.

192

The traditions that developed in the early days in the history of statistics in North Carolina continue to be evident as the concentration of statisticians in the state proliferates. The co- operative focus is not the only tradition. Another is the broad balance of theory and

application; another is the distinguished roster of faculty and graduates. Perhaps the healthiest characteristic has been not to become atrophied by tradition; statistics in North Carolina has developed with adaptability to changing times.

Notes and References

Adams, J. Stacy (1958). Interviewing Techniques. Chapel Hill: University of North Carolina Press. Cramer, H. and Leadbetter, M.R. (1967). Stationary and Related Stochastic Processes: Sample Function Pro-

perties and Their Applications. New York: John Wiley and Sons, Inc. Graham, F.P. (1948). 'A Unique Story', in the first Record of Research of the Institute of Statistics, page 4. Greenberg, B.G. (1959). Conduct of cooperative field and clinical trials. American Statistician, 13, 13-17, 28. Greenberg, B.G. (1964). Field training for biostatisticians. American Statistician, 18, 19-22. Institute of Statistics (1976). One activity of the Institute of Statistics that has continued is the publication of

its series of research reports. The most recent, Number 1080, was published in July 1976. Monroe, J. and Finkner, A.L. (1959). Handbook of Area Sampling. Philadelphia: Chilton Co. Sarhan, A.E. and Greenberg, B.G. (1962). Contributions to Order Statistics. New York: John Wiley and Sons,

Inc. Sheps, Cecil G. (chairman) (1976). Higher Education for Public Health: Report of the Milbank Memorial Fund

Commission, p. 122. New York: Prodist.

Appendix List of Exhibits for the History of Statistics in North Carolina

Department of Statistics (NCSU) Exhibit I. Current Faculty in Department of Statistics at North Carolina State

University (NCSU) as of end 1975-76 academic year Exhibit H. Former Faculty in Department of Statistics at North Carolina State

University (NCSU) as of end of 1975-76 academic year Exhibit III. Visiting Faculty in Department of Statistics at North Carolina State

University (NCSU) prior to end 1975-76 academic year Exhibit IV.1 Doctoral Degree Recipients from Department of Statistics at North Carolina

State University (NCSU). Exhibit V.' Master's Degrees Awarded by Department of Statistics at North Carolina

State University (NCSU).

Department of Statistics (UNC-CH) Exhibit VI. Current Faculty in Department of Statistics at University of North Carolina

at Chapel Hill (UNC-CH) as of end 1975-76 academic year. Exhibit VII. Former Faculty in Department of Statistics at University of North Carolina

at Chapel Hill (UNC-CH) prior to end 1975-76 academic year. Exhibit VIII. Visiting Faculty in Department of Statistics at University of North Carolina

at Chapel Hill (UNC-CH) prior to end 1975-76 academic year. Exhibit IX.' Doctoral Degree Recipients from Department of Statistics at University of

North Carolina at Chapel Hill (UNC-CH). Exhibit X.1 Master's Degrees Awarded by Department of Statistics at University of

North Carolina at Chapel Hill (UNC-CH).

1 Enquiries upon the detail in these Exhibits should be addressed to the Executive Editor. Alternatively, a complete copy may be inspected for reference at the Library of the Royal Statistical Society in London.

193

Department of Biostatistics Exhibit XI. Current Faculty in Department of Biostatistics at University of North

Carolina at Chapel Hill (UNC-CH) as of end 1975-76 academic year. Exhibit XII. Former Faculty in Department of Biostatistics at University of North

Carolina at Chapel Hill (UNC-CH) prior to end 1975-76 academic year. Exhibit XIII. Visiting Faculty in Department of Biostatistics at University of North

Carolina at Chapel Hill (UNC-CH) prior to end 1975-76 academic year. Exhibit XIV.' Doctoral Degree Recipients from Department of Biostatistics at University

of North Carolina at Chapel Hill (UNC-CH). Exhibit XV.' Master's Degrees Awarded by Department of Biostatistics at University of

North Carolina at Chapel Hill (UNC-CH). Exhibit I

Name Rank Date of appointment

Bhattacharyya, B.B. Professor September 1963 Cockerham, C.C. Professor September 1953 Cox, Gertrude M. Professor Emeritus November 1940 Gold, H.J. Professor July 1965 Grandage, A.H.E. Professor September 1952 Hader, R.J. Professor July 1948 Hafley, W.L. Professor June 1966 Hayne, D.W. Professor September 1962 Lucas, H.L. Professor January 1946 McVay, F.E. Professor August 1948 Mason, D.D. Professor September 1953 Monroe, R.J. Professor February 1941 Nelson, L.A. Professor March 1964 Proctor, C.H. Professor September 1960 Quesenberry, C.P. Professor September 1966 Rawlings, J.O. Professor July 1960 Ridgeway, D.L. Professor July 1967 Rigney, J.A. Professor January 1941 Steel, R.G.D. Professor September 1960 van der Vaart, H.R. Professor February 1962 Wesler, O. Professor July 1964 Gallant, A.R. Associate Professor July 1971 Gerig, T.M. Associate Professor September 1969 Giesbrecht, F.G. Associate Professor August 1971 Goodman, M.M. Associate Professor August 1967 Johnson, Thomas Associate Professor August 1974 Linnerud, A.C. Associate Professor July 1967 Manson, A.R. Associate Professor December 1965 Wasik, J.L. Associate Professor September 1967 Goodnight, J.H. Assistant Professor July 1970 Wakeley, J.T. Adjunct Professor April 1970 Bayless, D.L. Adjunct Associate Professor January 1969 Schreuder, H.T. Adjunct Assistant Professor December 1966

1 Enquiries upon the detail in these Exhibits should be addressed to the Executive Editor. Alternatively, a complete copy may be inspected for reference at the Library of The Royal Statistical Society in London.

194

Exhibit II

Name Terminal rank Years of employment

Anderson, Richard L. Professor 1942-67 Bartholomay, Anthony F. Professor 1967-69 Cochran, William G. Professor 1946-48 Comstock, Ralph E. Professor 1943-57 Cox, Gertrude M. Professor 1940-60 Finkner, Alva L. Professor 1941-60 Peach, Paul Professor 1946-50 Robinson, Harold F. Professor 1946-58 Shreve, Darrell R. Professor 1960-62 Smith, H. Fairfield Professor 1949-57 Stacy, Ralph W. Professor 1962-65 Wallace, T. Dudley Professor 1959-73 Williams, Evan J. Professor 1956-58 Hall, William J. (one-half time) Associate Professor 1960-63 Herbst, Laurence J., Jr. Associate Professor 1962-70 Horvitz, Daniel G. Associate Professor 1953-57 Koop, John C. Associate Professor 1960-67 Little, Charles H. Associate Professor 1968-70 Peterson, Roger G. Associate Professor 1962-65 Smart, William W.G., Jr. Associate Professor 1953-65 Teichman, Robert Associate Professor 1966-70 Warren, Jerry A. Associate Professor 1965-70 Burrows, Peter M. Assistant Professor 1971-75 Carter, Melvin W. Assistant Professor 1958-60 Cerimele, Benito J. Assistant Professor 1968-70 Chang, Monica Liu Assistant Professor 1967-69 Fleischer, Jack Assistant Professor 1953-60 Gray, Clifton W. (one-half time) Assistant Professor 1959-60 Greenwald, E.K. Assistant Professor 1966-69 Martin, D.C. Assistant Professor 1966-73 Matzinger, Dale F. Assistant Professor 1957-58 Meade, James H., Jr. Assistant Professor 1962-63 Mendenhall, William, III Assistant Professor 1958-59 Moll, Robert H. Assistant Professor 1957-58 Rohde, Charles H. Assistant Professor 1963-65 Williams, Mary B. Assistant Professor 1967-74 Carroll, Sarah P. Instructor 1947-56 Fleming, Margaret Instructor 1946-55 Foster, Walter D. Instructor 1949-50 Harris, Mary J. Instructor 1949-53 Teichroew, Dan Instructor 1950-53 Swindel, Benee F. Adjunct Associate Professor 1966-75

195

Exhibit IlI

Name Rank Dates

Box, G.E.P. Visiting Professor 1952-53 Carter, Melvin W. Visiting Associate Professor 1967, 1968, 1969 (summers only) Chapman, D.G. Visiting Professor 1958-59 Cooke, Dennis Visiting Associate Professor 1957-58 Cress, Charles E. Visiting Associate Professor 1973-74 Dixon, W.J. Visiting Professor 1951-52 Fryer, Holly C. Visiting Professor 1975 Gomez, F. Pimentel Visiting Associate Professor 1952-53 Griffiths, John Visiting Professor 1956 Kendall, M.G. Visiting Professor 1953 Kimura, Motoo Visiting Professor 1972 Li, Jerome C.R. Visiting Professor 1957 Morton, George Visiting Associate Professor 1956-57 Mott, G.O. Visiting Professor 1951-52 Mukai, Terumi Visiting Professor 1975, 1976 Nyquist, Wyman E. Visiting Professor 1969-70 Otha, Tomoko Visiting Assistant Professor 1972 Patterson, H.D. Visiting Associate Professor 1957-58 Pielou, Evelyn Visiting Professor 1967-68 Schultz, E.F., Jr. Visiting Assistant Professor 1953-54 Sen, A.R. Visiting Professor 1974 Shrikhande, S.S. Visiting Associate Professor 1960-61 Sisson, Donald V. Visiting Associate Professor 1975-76 Snedecor, G.W. Visiting Professor 1957-58 van der Vaart, H.R. Visiting Associate Professor 1957 Warren, John Visiting Instructor 1975-76 Watson, G.S. Visiting Professor 1959 Weir, Bruce S. Visiting Research Associate 1972, 1973, 1974 Wilkenson, Graham Visiting Professor 1962 Williams, E.J. Visiting Professor 1956-57 Wishart, John Visiting Professor 1949 Youden, W.J. Visiting Professor 1951

Special Summer Sessions

Bliss, C.I. Professor 1946 Bose, R.C. Professor 1951 Duncan, D.B. Professor 1956, 1959 Fisher, R.A. Professor 1946

Hotelling, Harold Professor 1941 Kossack, Carl F. Assistant Professor 1941 Nicholson, G.E. Professor 1959 Robson, D.S. Professor 1963 Roy, S.N. Professor 1951 Smith, W.L. Professor 1956, 1959 Snedecor, G.W. Professor 1941, 1946, 1951 Wolfowitz, J. Professor 1946 Youden, W.J. Professor 1951

196

Exhibit VI

Name Rank Date of appointment

Baker, Charles R. Professor 1968 Chakravarti, I.M. Professor 1964 Hoeffding, Wassily Professor 1947 Johnson, Norman L. Professor (Chairman 1971-76) 1962 Leadbetter, Malcolm R. Professor 1965 Simons, Gordon D. Professor (Chairman 1976-) 1968 Smith, Walter L. Professor 1954 Cambanis, Stamatis Associate Professor 1969 Kelly, Douglas G. Associate Professor 1968 Smith, K.J.C. Associate Professor 1968 Wegman, Edward J. Associate Professor 1968 Carroll, Raymond J. Assistant Professor 1974

Exhibit VII

Name Terminal rank Years of employment

Bose, Raj Chandra Kenan Professor Emeritus 1947-71 David, Herbert A. Professor 1968-69 Duncan, David B. Professor 1955-61 Hall, William J. Professor 1957-68 Hotelling, Harold Kenan Professor Emeritus 1964-died 1973 Nicholson, George E., Jr. Professor 1947-died 1971 Robbins, Herbert E. Professor 1945-53 Roy, Samarendra Nath Professor 1948-died 1964 Gould, Floyd J. Associate Professor 1968-73 Hsu, P.L. Associate Professor 1946-48 Madow, William G. Associate Professor 1946-48 Bohrer, Robert E. Assistant Professor 1965-67 Cleveland, William S. Assistant Professor 1969-71 Dowling, Thomas Allan Assistant Professor 1967-December 1971 Shachtman, Richard H. Assistant Professor 1968-73 Smith, Woollcott K. Assistant Professor 1970-January 1972 Paulson, Edward Instructor 1946-47

Exhibit VIII

Name Dates

Adams, John William 1962-65 Aigner, Martin Baldessari, Bruno A. September 1965-June 1966 Barlotti, Adriano September 1964-June 1966 Bartlett, Maurice Stevenson 1945-46 Basu, Debabrata August 1964-July 1965 Behboodian, Javad August 1970-May 1971 Bentley, Donald L. August 1970-August 1971 Berlekamp, Elwyn R. September 1966-December 1966 Bertaud, Marcel 1962-63

197

Exhibit VIII (continued)

Name Dates

Bhapkar, Vasant P. 1964-66 Bhattacharyya, Helen August 1975-May 1976 Bhattacharyya, P.K. 1961-62 Bogdanoff, John L. September 1965-June 1966 Bruck, R.H. September 1963-May 1964 Camion, Paul January 1974-May 1974 Chu, John T. 1951-52, 1953-55 Clatworthy, W.H. September 1968-June 1969 Cohan, Avery 1961 Corsten, Leo Caspar Antoon February 1958-June 1959 Dall'Aglio, Giorgio September 1962-June 1963 Deshpande, Jayant Vinayak September 1966-August 1967 DeWaal, Daniel J. August 1974-May 1975 Diananda, P.H. September 1953-July 1954 Durbin, James September 1958-May 1959 Eicker, Friedheim February 1959-September 1960 Engler, Jean Ann September 1957-August 1958 Fand, David I. June 1958-59 Foata, Dominique Gyprien August 1974-November 1974 Freeman, Harold July 1959-June 1961 Gabriel, Kuno Ruben September 1957-July 1958 Gere, B.H. September 1970-May 1971 Gupta, Ram Prakash September 1967-July 1968 Guthrie, Donald, Jr. August 1969-June 1970 Hannan, Edward Joseph October 1958-May 1959

Hayakawa, Takesi September 1969-August 1970 Heathcote, Christopher September 1961-August 1962 Hunter, Jefferey Joseph October 1968-January 1969; January 1973-June 1973

Ibragimov, Ildar A. October 1973-February 1974 Ikeda, Sadao 1965-67 Ito, Paul K. 1954-55 Joshi, V.M. June 1966-September 1966

Kallenberg, Olav September 1973-May 1974 Katz, Leo Summer of 1950 Khatri, Chinubhai Ghelabhai February 1964-January 1965

King, Randall M. August 1974-May 1975

Kojima, T. 1971-72 Kotz, Samuel September 1962-June 1963

Kozelka, Robert M. September 1963-June 1964 Krafft, Olaf September 1963-June 1964

Krier, Nicholas September 1970-June 1971

Kulldorf, Gunnar K.O. 1952-53 Laskar, Mrs Renu Chakravarti 1967-68 Lee, Doris Margaret September 1951-May 1952

Linder, Arthur September 1963-May 1964 Lindgren, Sven George Kristian September 1970-May 1971 Marsaglia, George 1953-54

198

Exhibit VIII (continued)

Name Dates

Masry, Elias August 1973-May 1974 Masuyama, Motosaburo September 1965-May 1966 McGilchrist, Clyde A. January 1974-June 1974 Mesner, Dale Marsh September 1964-May 1965 Miller, Hilton D. September 1964-May 1965 Momoi, Makoto August 1968-69 Morimura, Hidenori September 1968-May 1969 Moustafa, Madany D. November 1956-September 1957 Nakamura, Akira September 1966-June 1967 Naor, Pinhas Paul September 1962-June 1964; July 1966-October 1966 Nemenyi, Peter Bjorn September 1965-June 1966 Ogawa, Junjiro 1955-65 Page, Ewan Stafford July 1962-April 1963 Paulson, Edward September 1947-June 1948 Philipp, Walter August 1972-May 1973 Pitman, Edward James George 1948-49 Potthoff, Richard F. 1965-66 Quails, Clifford 1970-71 Raghavarao, D. August 1972-May 1973 Rajput, Balram S. August 1969-June 1972 Ray-Chaudhuri, Dwijendra K. 1960-61 Robertson, Tim September 1974-May 1975 Rootzen, Holger September 1975-June 1976 Roy, Jogabrata October 1959-May 1960 Ruist, Erik Harald 1948-49 Runnenburg, Johannes September 1964-August 1966 Saw, John Grenville September 1959-August 1960 Schutzenberger, Marcel P. September 1960-August 1961 Sethuraman, Jayaram September 1962-May 1963 Shrikhande, S.S. 1951-52; 1958-59 Siotani, Minoru August 1968-September 1969 Spivey, Walter Allen 1956-57 Srivastava, Jagdish Narain 1961-63 Stout, William F. August 1974-May 1975 Sugiyama, Takakazu February 1967-September 1968 Suigiura, Nariaki September 1967-August 1968 Sundrum, Ramaswamy Meenatchi July 1953-June 1954 Tate, Robert F. 1948-49 Tranquilli, Giovanni B. September 1964-April 1965 Vuagnat, Pierre September 1971-December 1971 Watanabe, Hisao 1970-71 Watson, G.S. 1957-58 Zyskind, George 1958-59

199 Exhibit XI

Name Current faculty rank Initial appointment

Abernathy, James R. Professor 1 July 1965 Elandt-Johnson, Regina C. Professor 1 September 1964 Elston, Robert C. Professor 1 July 1960 Greenberg, Bernard G. Dean of the School 1 July 1949

of Public Health and Kenan Professor

Grizzle, James E. Department 1 January 1960 Chairman and Professor

Koch, Gary G. Professor 1 December 1967 Kuebler, Roy R., Jr. Professor 1 September 1958-30 June 1976 Lachenbruch, Peter A. Professor 1 June 1965-30 June 1976 Linder, Forrest E. Professor 1 September 1966 Quade, Dana E. A. Professor 1 August 1962 Sen, Pranab K. Professor 1 August 1965 Wells, Henry Bradley Professor 1 February 1958 Coulter, Elizabeth J. Professor 1 September 1965 Davis, Clarence E. Associate Professor 1 June 1972 Gillings, Dennis B. Associate Professor 1 May 1971 Helms, Ronald W. Associate Professor 1 August 1968 Kleinbaum, David G. Associate Professor 1 July 1970 Kupper, Lawrence L. Associate Professor 1 July 1970

Lingner, Joan W. Associate Professor 1 December 1970 Shachtman, Richard H. Associate Professor 1 June 1973 Symons, Michael J. Associate Professor 1 August 1969 Williams, 0. Dale Associate Professor 1 October 1970 Bilsborrow, Richard E. Assistant Professor 15 October 1972 Francis, Mildred E. Assistant Professor 1 August 1974 Hogue, Carol R. Assistant Professor 1 September 1974

Imrey, Peter B. Assistant Professor 1 July 1972-31 July 1975 Rademaker, Alfred W. Assistant Professor 15 June 1974 Smith, Wendell C. Assistant Professor 1 July 1974 Stewart, J. Richard Assistant Professor 1 September 1970-30 June 1976 Suchindran, Chirayath M. Assistant Professor 1 September 1972 Turnbull, Craig D. Assistant Professor 1 August 1971

Kaplan, Ellen B. Instructor 1 September 1959 Sorant, Alexa M. Instructor 1 September 1970-30 November 1976

Grimson, Roger C. Research Associate 1 September 1976 Professor

Adlakha, Arjun L. Research Assistant 1 July 1975 Professor

Ahmed, Susan W. Research Assistant 1 January 1976 Professor

Samuels, Steven J. Research Instructor 1 March 1976 Baker, Raymond H. Lecturer and 1 November 1973

currently Associate Director for Operations at the Population Lab

200

Exhibit XI (continued)

Name Current faculty rank Initial appointment

Blake, Martha Ellan Lecturer 1 March 1974 Chanlett, Eliska Research Associate 1 March 1970 Krusa, Hans Research Associate 1 January 1974 Namboodiri, Kadambari Research Associate 1 July 1969

Exhibit XII

Name Terminal rank Years of employment

Bartholomay, Anthony F. Professor 1 September 1967-30 June 1969 David, Herbert A. Professor 1 September 1964-31 August 1972 Sheps, Mindel C. Professor 1 December 1968-15 January 1973 Smith, Harry, Jr. Instructor 1 July 1953-30 June 1954

Assistant Professor 1 July 1954-30 June 1955 Professor 1 July 1966-30 June 1971

Donnelly, Thomas G. Associate Professor 1 September 1956-30 June 1971 Kosa, John Associate Professor 1 December 1960-29 February 1964 Brogan, Donna Ruhl Assistant Professor 1 August 1967-31 August 1971 Cockerham, C. Clark Assistant Professor 1 July 1952-30 June 1953 Diamond, Earl L. Assistant Professor 1 September 1957-1 July 1959 Glasser, Jay H. Assistant Professor 9 September 1964-31 August 1969 Hassanein, Khatab M. Assistant Professor 1 July 1962-31 July 1964 Pasternack, Bernard S. Assistant Professor 1 October 1959-31 January 1961 Cranford, Sarah Instructor 1 July 1970-June 1973 Fergany, Nader Instructor and 1 February 1970-30 June 1971

Visiting Assistant Professor

Gehan, Edmund A. Instructor 1 July 1955-31 January 1958 Johnson, William D. Instructor 1 August 1968-30 June 1974 Leiva, Miriam A. Instructor 1 July 1966-30 November 1966 Ridgway, Linda S. Instructor 1 September 1968-28 February 1970 Stephenson, Ruth E. Instructor 1 July 1964-31 December 1964 Yelverton, Katie C. Instructor 1 December 1966-15 August 1972 Sarhan, Ahmed E. Research Associate 1 July 1955-31 August 1956

Research Professor 1 September 1956-15 August 1970 and Visiting (intermittent employment) Professor

Back, Kurt W. Research Associate 1 September 1956-31 August 1959 Professor

Miller, Norman Research Associate 15 September 1958-1 September 1961 Professor

Phillips, Bernard S. Research Assistant 1 September 1956-31 August 1958 Professor

Strickland, Lloyd H. Research Assistant 1 September 1959-1 July 1960 Professor

Lieberman, Milton D. Lecturer 1 February 1973-30 June 1974

201

Exhibit XIII

Name Rank and home institution Dates

Bilsborrow, Richard E. Visiting Assistant Professor 15 October 1972- Assistant Professor at New York 15 October 1973

University, Department of Economics, New York City, New York

Chakravorti, Sukharanjan Visiting Assistant Professor 1 January 1973- Lecturer in Statistics at Burdwan 1 June 1974

University, Department of Economics, West Bengal, India

Chatterjee, Shoutir K. Visiting Associate Professor 1 January 1972- Associate Professor of Statistics at 31 December 1972

Calcutta University, Calcutta, India Chiang, Chin Long Visiting Professor 17 June 1963-

Associate Professor of Biostatistics at 31 August 1970 the University of California, School (intermittent of Public Health, Berkeley, employment) California

Cooke, Dorothy S. Visiting Professor 1 November 1969- Chief, International Research Branch, 30 April 1970

Office of International Statistical 1 May 1970- Programs, National Center for 30 June 1974 Health Statistics, U.S. Public (Adjunct Professor) Health Service, Rockville, Maryland

Council, Charles R. Visiting Associate Professor 1 July 1962- Chief of the Public Health Statistics 30 September 1962

Section, North Carolina State Board of Health, Raleigh, North Carolina

Cox, David R. Visiting Professor 1 August 1955- Professor at Birkbeck College, 31 July 1956

University of London, Department of Statistics, London, England

Davis, Arthur W. Visiting Associate Professor 1 July 1969- Senior Research Scientist at the 30 June 1970

University of Melbourne, Division of Mathematical Statistics, Melbourne, Australia

Donnelly, Thomas G. Visiting Associate Professor 1 January 1971- Private Research Consultant, 30 June 1971

associated with a large number of university projects, University of North Carolina, Chapel Hill, North Carolina

Dowd, John E. Visiting Instructor 15 September 1970- Statistician, Mathematics Statistics 31 January 1971

Unit Division of Research in Epidemiology, World Health Organization, Geneva, Switzerland

202

Exhibit XIII (continued)

Name Rank and home institution Dates

Duncan, David B. Visiting Professor 2 February 1958- Professor, University of Florida, 31 August 1958

Department of Statistics, Gainesville, Florida

Edelman, David A. Visiting Assistant Professor 1 July 1972- Graduate of the University of North 30 June 1974

Carolina at Chapel Hill in the Department of Biostatistics

Elandt-Johnson, Regina C. Visiting Associate Professor 1 September 1964- Research Assistant Lecturer and Head 1 September 1965

of the Statistics Department at the Agricultural University, Department of Agricultural Experimentation, Poznan, Poland

Fergany, Nader A. Visiting Assistant Professor 1 July 1970- Instructor, Cairo University, 30 June 1971

Department of Statistics, Cairo, Egypt

Flinchum, Glenn A. Visiting Assistant Professor 1 October 1962- Chief, Public Health Statistics 31 March 1965

Section, North Carolina State (intermittent Board of Health, Raleigh, North employment) Carolina

Fryer, John G. Visiting Associate Professor 1 August 1971- Lecturer in Statistics, Exeter University, 30 June 1972

Exeter, Devon, England Gabriel, K. Ruben Visiting Associate Professor 1 September 1966-

Associate Professor of Statistics at 31 August 1967 Hebrew University, Jerusalem, Israel

Gamrah, Hamed A. Visiting Assistant Professor 1 August 1975- Graduate in demography at the 31 December 1975

Catholic University of Louvain, Belgium

Ghosh, Manindra Visiting Professor 1 January 1953- Lecturer at the University of Calcutta, 1 July 1954

Department of Statistics, Calcutta, India

Goodman, J. Roe Visiting Professor 1 July 1972- Statistical Advisor in various countries 31 August 1973

at the United Nations and affiliated agencies

Gower, John C. Visiting Associate Professor 1 September 1970- Principal Scientific Officer, Rothamsted 31 August 1971

Experimental Station, Harpenden, Herts, England

203

Exhibit XIII (continued)

Name Rank and home institution Dates

Hamilton, C. Horace Visiting Research Professor 1 February 1967- Professor of Rural Sociology at the 30 June 1971

North Carolina State College, Raleigh, North Carolina

Imrey, Peter B. Visiting Assistant Professor 1 July 1972- Research Assistant at the University of 1 July 1974

North Carolina at Chapel Hill, Department of Biostatistics

Irwin, J.O. Visiting Professor 1 July 1958- Lecturer, London School of Hygiene, 30 June 1962

University of London, London, (intermittent W.C.1., England employment) 1

January 1965- 30 June 1965

John, Kallacheril S. Visiting Assistant Professor 1 June 1968- Senior Lecturer at the Australian 31 January 1969

National University, Canberra, Australia

Joshi, Prakash C. Visiting Assistant Professor 1 October 1969- 31 August 1970

Krotki, Karol J. Visiting Professor 1 October 1972- Professor of Sociology at the 30 June 1974

University of Alberta, Edmonton, Canada

Laby, Betty Adjunct Professor 1 May 1967- Research Assistant at the University 24 July 1967

of Melbourne, Melbourne, Australia

Linder, Forrest E. Visiting Professor 1 September 1966- Director of the National Center for 31 January 1967

Health Statistics, Rockville, Maryland Losee, Garrie J. Adjunct Professor 1 September 1972-

Mathematical Statistician in the 30 June 1974 Office of the Director, Bureau of Health Professions, Education and Manpower Training, National Institutes of Health, Bethesda, Maryland

McArthur, Norma Visiting Professor 15 October 1968- Professional Fellow - acted as the 30 November 1968

consultant/advisor for the Australian National University, Canberra, Australia

McGilchrist, Clyde A. Visiting Associate Professor 1 July 1974- Associate Professor in Statistics at the 31 December 1974

University of New South Wales, Sydney, Australia

204

Exhibit XIII (continued)

Name Rank and home institution Dates

Myers, Robert J. Visiting Professor 1 March 1970- Manpower Advisor to the government 31 December 1973

of Zambia, Lusaka through the Ford Foundation

Nathan, Gad Visiting Associate Professor 1 August 1971- Senior Teacher at the Tel Aviv 31 July 1972

University, Department of Statistics, Tel Aviv, Israel

Newell, David J. Visiting Professor 1 July 1959- Senior Lecturer in Biostatistics 1 August 1960

at the Medical School, Newcastle upon Tyne, England

Ogawa, Junjiro Visiting Associate Professor 1 September 1958- Research Associate in the Institute of 1 July 1959

Statistics at the University of North Carolina at Chapel Hill

Puri, Madan L. Visiting Associate Professor 15 May 1966- Associate Professor at the New York 15 September 1967

University, Courant Institute of Mathematical Sciences, New York City, New York

Read, Kenneth L. Visiting Assistant Professor 1 August 1975- Lecturer at the University of Exeter, 31 July 1976

Department of Mathematical Statistics and Operational Research, Exeter, EX4 4PU, Devon, England

Rose, J. Hugh Visiting Professor 1 February 1972- Statistical Advisor to Kenya, U.S. 30 June 1974

Bureau of the Census, Washington, D.C.

Sabagh, Georges Adjunct Professor 1 September 1969- Chairman, at the University of 30 June 1975

California, Department of Sociology, Los Angeles, California

Sarndal, Carl-Erik Visiting Assistant Professor 1 September 1962- Research Associate at the University 30 June 1963

of North Carolina at Chapel Hill, Department of Biostatistics

Scott, Alastair J. Visiting Assistant Professor 1 September 1969- Lecturer in Statistics at the University 31 August 1970

of London, The London School of Economics, London, England

Sen, Pranab K. Visiting Associate Professor 1 August 1965- Visiting Assistant Professor of 1 July 1966

Statistics at the University of California, Berkeley, California

205

Exhibit XIII (continued)

Name Rank and home institution Dates

Siervogel, Roger M. Visiting Assistant Professor 1 August 1973- Research Associate at the University 31 October 1974

of North Carolina at Chapel Hill, Department of Biostatistics, Genetic Laboratory

Suchindran, Chirayath M. Visiting Assistant Professor 1 September 1972- Graduate of the University of North 1 September 1973

Carolina at Chapel Hill, Department of Biostatistics

Thomlinson, Ralph Visiting Professor 1 July 1972- Professor and Chairman at California 30 June 1973

State College, Department of Sociology, Los Angeles, California

Special Summer Session

Abbey, Helen Visiting Associate Professor 27 June 1964- Associate Professor, Department of 7 August 1964

Biostatistics, School of Hygiene and Public Health, Johns Hopkins University, Baltimore, Maryland

Anderson, Paul S. Visiting Assistant Professor 17 June 1963- Assistant Professor in Public Health 26 July 1963

(Biometry), Department of 29 June 1964-

Epidemiology, Yale School of 7 August 1964 Medicine, New Haven, Connecticut

Bailey, Albert E. Visiting Professor 29 June 1964- Director, Bureau of Administration 7 August 1964

and Management, Pennsylvania Department of Health, Harrisburg, Pennsylvania

Brown, Byron Wm., Jr. Visiting Associate Professor 29 June 1964- Associate Professor, Department of 7 August 1964

Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota

Chiang, Chin Long Visiting Associate Professor 17 June 1963- Associate Professor of Biostatistics, 26 July 1963

School of Public Health, University of California, Berkeley, California

Colton, Theodore Visiting Professor 29 June 1964- Research Associate, Department of 7 August 1964

Preventive Medicine, Medical School, Harvard University, Boston, Massachusetts

46/2-F

206

Exhibit XIII (continued)

Name Rank and home institution Dates

Cornell, Richard G. Visiting Associate Professor 29 June 1964- Associate Professor of Statistics and 7 August 1964

Director of the Biometry Training Program, Florida State University, Tallahassee, Florida

David, Herbert A. Visiting Professor 17 June 1963- Professor of Statistics, Virginia 26 July 1963

Polytechnic Institute, Blacksburg, Virginia

Elston, Robert C. Visiting Associate Professor 29 June 1964- Senior Research Fellow, Department 7 August 1964

of Statistics, University of Aberdeen, Old Aberdeen, Scotland

Erhardt, Carl L. Visiting Professor 17 June 1963- Director, Bureau of Records and 26 July 1963

Statistics, City of New York, Department of Health, New York City, New York

Hemphill, Fay M. Visiting Professor 17 June 1963- Assistant Chief, Division of Research 26 July 1963

Grants (1963) and Chief, Scientific 29 June 1964- and Technical Information Office, 7 August 1964 National Cancer Institute (1964), National Institutes of Health, Bethesda, Maryland

Horvitz, Daniel G. Visiting Professor 17 June 1963- Head, Sampling Group, Statistics 26 July 1963

Research Division, Research 29 June 1964- Triangle Institute, Research 7 August 1964 Triangle Park, North Carolina

Johnson, Eugene A. Visiting Professor 17 June 1963- Professor, Gustavus Adolphus 26 July 1963

College, St Peter, Minnesota

Li, C.C. Visiting Professor 17 June 1963- Professor of Biometry, School of 26 July 1963

Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania

McMahan, C.A. Visiting Professor 17 June 1963- Head, Department of Biostatistics, 26 July 1963

School of Medicine, Louisiana State University, New Orleans, Louisiana

Newell, David J. Visiting Professor 17 June 1963- Senior Lecturer in Biostatistics, The 26 July 1963

Medical School, Newcastle upon Tyne, England

207

Exhibit XIII (continued)

Name Rank and home institution Dates

Perrin, E.B. Visiting Assistant Professor 29 June 1964- Assistant Professor, Department of 7 August 1964

Preventive Medicine, School of Medicine, University of Washington, Seattle, Washington

Remington, Richard D. Visiting Associate Professor 17 June 1963- Associate Professor of Biostatistics, 26 July 1963 School of Public Health, 29 June 1964-

University of Michigan, Ann Arbor, 7 August 1964 Michigan

Tweedie, M.C.K. Visiting Assistant Professor 29 June 1964- Lecturer in Mathematical Statistics, 7 August 1964

The University of Liverpool, Liverpool, England

White, Colin Visiting Professor 29 June 1964- Professor of Biometry, Department 7 August 1964

of Epidemiology and Public Health, Yale School of Medicine, New Haven, Connecticut

Resume

L'expansion que la statistique en tant que discipline universitaire a prise dans le syst6me universitaire de Caroline du Nord est n6e avant Pearl Harbor. La croissance initiale - qui ffit frein6e de certains c6t6s mais aussi stimul6e pour d'autres durant les ann6es de la Seconde Guerre Mondiale - a constitu6 une base de depart solide pour la croissance accl16r6e survenue apres guerre, tout en pr6servant la haute qualit6 des travaux et leur ad6quation aux besoins de l'6tat, de la region et de la nation, caract6ristiques de l'expansion de l'6poque pr6- sente.

Ce qu'on appelle maintenant le d6partement de statistique (Ecole des sciences math6matiques et physiques de l'Universit6 d'Etat de Caroline du Nord) fi1t fond6 Alafinde 1940. Ce d6partement de pointe connut une demande rapidement croissante (au plan local et national) en matiere de formation, de participation A la recherche et d'aide sous forme de consultations. Pour aider "A faire face A ces besoins, deux autres d6partements de statistique furent cre6s rl'Universit6 de Caroline du Nord, a Chapel Hill, lesquels sont maintenant le d6partement de Statistique de l'Ecole des Arts et Sciences, (cre6 en 1946) et le d6partement de Biostatistique de l'Ecole de Sant6 Publique (cr66 en 1949). Tous les d6veloppements du debut en mati"re de statistique recurent I'appui 6nergique de Frank Porter Graham, President du systhme universitaire de Caroline du Nord qui incluait alors trois institutions de base, A Chapel Hill, A Raleigh et a Greensboro. L'entit6 connue sous le nom d'Institut de Statistique fi1t constitu6e A Raleigh en 1944 et a requ en 1946 son statut (systeme universitaire de Caroline du Nord).