UF-Statistics UFL

The Department of Statistics, 1963-1977:

The Bird That Almost Did Not Fly

by William Mendenhall III

The following article was prepared by Dr. William Mendenhall, first chair of the Statistics Department, for the 25th anniversary of the founding of the department in 1988.
The first six years of the Department of Statistics, 1963 through 1968, presented an unending sequence of emotional valleys and peaks, the valleys corresponding to seemingly catastrophic roadblocks to our progress that, in my naivete, I failed to anticipate, the peaks to their eventual solutions.

It has been noted that, at the time when the Department of Statistics was formed, course offerings in statistics were scattered across at least eight departments on campus. Not only did this condition exist when the department was formed, it existed more than a year later when I arrived at the University of Florida to be interviewed for the chairmanship. There were no faculty members in the so-called department (other than Dr. Brandt who served as the Experiment Station consulting statistician), no vacant position line items, no budget, no approved degrees and no courses.

I arrived for the interview late on a balmy evening in March 1963 and checked into the old Holiday Inn on south 13th Street. The motel was teeming with high school students, boys and girls shouting, screaming, and running from room to room. Unknowingly I had timed my visit to coincide with the Florida high school basketball championships.

After a sleepless night, I met Dr. Brandt, an early riser, at 7:00 AM for breakfast. Suffering from a violent headache, I listened as he described the living conditions in Gainesville. Then I asked him about the University's plan for the new Department of Statistics. He aggravated my headache when he told me that the sole person to be hired was to be the new chairman. The remainder of the departmental staff members were to be drawn from the various departments in the Colleges Agriculture, Arts and Sciences, Business and Engineering. Only one of these persons had a graduate degree in statistics, Dr. Willard Ash, who at that time was teaching courses in statistical methods an the design of experiments in the Department of Agronomy. I can distinctly remember my feeling utter disgust and thinking that I would like to leave immediately for Jacksonville and take the first flight back to Pennsylvania.

I met with many University of Florida faculty members and administrators during my two day visit, among them J. Wayne Reitz, President of the University, Robert Mautz, Vice-President for Academic Affairs, Joe Beckenbach, Director of the Agricultural Experiment Station, Marvin Brooker, Dean of the College of Agriculture and Ralph Page, Dean of the College of Arts and Sciences. I had known Ralph Page as Dean of Men at Bucknell University when I attended Bucknell as a freshman in 1943. These meetings created a vague picture of the University's plan for the new department.

Courses for the department were initially to be acquired by transferring statistics courses currently taught in the Departments of Mathematics and Agronomy, the College of Business, etc., to the new Department of Statistics. It seemed fairly clear to me that the Department of Mathematics wanted to retain control of its statistics courses and the Ph.D. program in statistics which it had historically purported to offer as a specialty within the mathematics Ph.D. program. Other departments teaching statistics courses as part of their curriculum may have feared the possibility of losing control of their courses but were comforted by the expectation that their own faculty members would move with the courses to the new department. I felt fairly certain that resistance to a transfer of courses would stiffen if the faculty for the new statistics department were to consist solely of newly hired Ph.D. statisticians.

On the positive side, two strong forces favored the formation of the new Department of Statistics. Agriculture, specifically the College of Agriculture and the Agricultural Experiment Station, backed by Dr. Reitz, were in desperate need of one or more statisticians to serve as statistical consultants upon Dr. Brandt's impending retirement on July 1, 1963. With the national scarcity of statisticians, administrators in agriculture were beginning to realize that a new department might be necessary in order to attract qualified statistical consultants. The second strong force behind the creation of a new department of statistics was Robert Mautz, Vice President for Academic Affairs Faced with a tight university budget, he saw the opportunity of saving money by combining the many small section introductory courses taught throughout the university into a few large section course taught by a new department of statistics.

Vice President Mautz was the key to forming a new department staffed by professional statisticians, because the teaching line items (with the exception of the one occupied by Willard Ash) would have to come from his budget. My last meeting at the University was with Dr. Mautz. I can remember telling him that he would never form a good department of statistics unless he staffed it with a core of Ph.D. statisticians and I can also remember his telling me that he would never do it. Ten days later I received a telephone call offering me the chairmanship with a budget containing five line items to be staffed solely by professional statisticians. I was told that Bob Mautz, the person most opposed to my suggestions for staffing the proposed department, was its strongest supporter.

Approval of a new department with a total of six line items seemed the ultimate victory. As I was about to learn, our difficulties were just beginning. I did not anticipate the barriers that we would encounter as a result of university bureaucracy and politics.

Assembling the elements of a department, faculty, courses and students is a multicircular problem in the bureaucratic setting of a university. It is very difficult to attract qualified faculty until you have courses, students and degree programs. And you cannot obtain approval for new courses and degree programs until you present adequate proof of student demand for them. And, of course, you cannot attract students to your courses if you have neither courses nor faculty to teach them. New courses must be approved by college curriculum committees and new graduate programs must approved by both the Graduate Council and the Board of Regents. It can take one or many month before a proposal is placed on a committee agenda and opposition by interested parties can and do delay approval of a proposal or result in its rejection.

The immediate problem in the spring of 1963 was solving the staffing problem for the coming September. We had no idea how many students we would inherit along with the courses transferred to the department but it seemed likely that the number of required sections would be more than Willard Ash and I could teach.

Our first stroke of luck, perhaps the key to our survival, was my meeting Larry Kupper at short courses in statistics that I was teaching for the U.S. Bureau of Mines in 1962 and 1963. Larry was an engineer and was head and shoulders above other course attendees, masters and Ph.D.'s, in comprehension of the concepts of statistics. I was convinced that he was an excellent prospect for graduate study in statistics and that he possessed the teaching ability that was needed to assist in teaching our new undergraduate level introductory courses. As it turned out, we inadvertently deceived each other. I recruited him for a master's program in statistics which I mistakenly thought would be approved late in 1963. He failed to tell me that he had a C undergraduate average and therefore was ineligible for admission to the Graduate School. After much argument with the Graduate School and the Admissions Office, I obtained his probationary admission to the Graduate School. This effort we repaid many times. He was an excellent graduate student and an outstanding teacher. He was to become the indispensable backbone for teaching our large section introductory courses.

Larry was joined in the fall by a hardworking graduate in mathematics named John Cornell. John and Larry were the first two graduate students in the Department of Statistics. Larry is currently a Professor of Biostatistics at the University of North Carolina. John is a Professor of Statistics at the University of Florida.

Arriving at Gainesville in the fall of 1963, I found that the Mathematics Department was refusing to transfer their courses in statistics, except for the freshmen introductory courses (which they did not want to teach), to the Department of Statistics. The Department ofAgricultural Economic refused to transfer its course on sample survey design and several departments were refusing to eliminate their competing introductory courses. (Robert Mautz spurred these transfers!) In addition, I learned that the process necessary to receive approval for a master's program in statistics was long and tedious. Dean L.E. Grinter, who had seemed so supportive when I visited for interviews, was readying himself and the Graduate School to put us through the hoops. And so, he did. We encountered months of tedious delays to approval of our proposed graduate programs.

Then we had to submit our proposal for the master's program to the Board of Regents. I can remember receiving a copy of a letter from the President of Florida State University to President Reitz which noted the applied skills possessed by our faculty and the theoretical interests and backgrounds of the F.S.U. faculty. He therefore suggested that the University of Florida teach only graduate courses in applied statistics and F.S.U. would teach only courses in statistical theory. Needless to say, we avoided that torpedo.

Recruiting faculty for the four vacant faculty positions was a problem. One well known U.S. statistician is said to have given us less than a chance of a snowball in - well, it wasn't put exactly that way - but, the idea was that we did not have a very good chance of hiring any faculty in the existing tight market for statisticians. I was convinced that one or several of my friends or contacts would accept the positions and the relatively high salaries attached to them. Six months and 48 contacts later, first by letters and later by phone calls, I was still unsuccessful. I was constructing a small list of prospects by examining publications in the major statistical journals. This is how we had our second bit of luck. I encountered the name of P.V. Rao. After an interview, I was convinced that he had the potential for research and that he would be a good teacher.  Unknown to me was the fact that P.V. and his wife, Premila, planned to return to India within a year.

Fred Barnett, a John Saw Ph.D. graduate at V.P.I., accepted our second teaching appointment. Thus, we started 1964 with Willard Ash, P.V. Rao, Fred Barnett, Frank Martin and myself. Frank Martin accepted the position of Associate Statistician, i.e., as a consultant, in the Agricultural Experiment Station. We were joined by some graduate students, two of them from Bucknell University, Robert Beaver and Richard Scheaffer.

The third stroke of luck occurred in 1965 when we added John Saw to our faculty. John was and is an outstanding statistician, an excellent teacher and a solver of problems. His addition to the department provided the core that we needed to form a viable department of statistics.

This was the beginning. To this core of faculty, we added K.C. Chanda, Lyman Ott and Dick Scheaffer. Then came Alan Agresti, Jay Devore, Ramon Littell, Jim McClave, Ron Marks, Jonathan Shuster, Dennis Wackerly and Mark Yang.

The high points of those years were approvals of our master and Ph.D. degree programs, events that occurred in 1965 and 1968, respectively.

Looking back on it now, it seems as though it happened so automatically, that it was so easy. But I can remember months of delays as we sought the approval of the Graduate Council for both our masters and Ph.D. degree programs. I also remember the opposition to our Ph.D. program proposal that we encountered within the state university system. In fact, it was not clear for several years whether our Ph.D. program would be approved.

In the meantime, we had to attract graduate students on faith. Thus we partially owe our existence today to Paul Benson of Bucknell University for his faith in our program and for providing the only single reliable source of graduate students during those lean years. We never would have made it without the support of Bob Mautz who gave us line items as they were needed and helped us against our internal opposition. And, needless to say, we never would have been able to survive had it not been for the graduate students and faculty who stayed the course despite our many setbacks.

To all of you, I offer my congratulations and thanks on this 25th anniversary of the birth of the Department of Statistics at the University of Florida.


Information Academic Programs Personnel
Departmental Units [an error occurred while processing this directive]Department Resources Links