“Gertrude was a leader; her concern was building an institution of statistics.”

I recall her being totally unassuming. She interacted with students in a most positive way. I vividly remember her advice to students to learn a substantive field to accompany the statistical courses. This was a time when decision theory was a hot area and mathematization was highly regarded. Many of us did not take her advice. Her advice resonated with me 10 years later, when I had a joint appointment between statistics and education. The education part was a catalyst for some of my multivariate papers and the origin of my interest in meta-analysis.

Gertrude was a leader; her concern was building an institute of statistics and not her own advancement.

She chose Harold Hotelling as chair at Chapel Hill, and she was chair at NCSU. She also chose Bernie Greenberg to be chair of biostatistics.

There is a story about Gertrude, whose truth is unclear. She wanted to hire Hotelling, who was at Columbia at that time, but she was unsure whether he would even consider an offer. She had been friendly with Susannah Hotelling and discussed this with her. Susannah suggested she just ask him with the question, “What would it take for you to come to Chapel Hill as chair?” She did and Hotelling’s reply was, “A salary higher than the football coach.” This seems odd for someone of Hotelling’s stature, except that the football coach was the highest-paid member at UNC, and this may have been a way for Hotelling to assert that scholarship should dominate football. As an aside, let me note that football was big at UNC when I was a student. But there was a catch. A salary higher than the coach had to be approved by the governor. So we are back to Gertrude, who had to convince the governor to approve, which she did.

During my days at UNC, we took some classes at NCSU. I took a class from Bill Cochran. Later on, the NCSU faculty lectured on the UNC campus. Gertrude visited, and I was always aware of her presence. She was a major champion of statistics.”

—Ingram Olkin, PhD ’51

“She really wanted to go to Washington to present her own testimony, but her physician would not permit it. She was angry and frustrated.”

In March of 1950, my GI Bill funding ran out just as I completed my master’s degree requirements in the department of statistics in Raleigh. Jack Rigney, chair of the department, asked me to present the results of my work at the next department seminar. As was customary, I selected a humorous article from the current Saturday Evening Post to introduce my presentation.

It was the story of a brief encounter between one of a community’s highly respected preachers and the chair of the local school system, who had just arrived with a doctorate from The Teachers’ College at Columbia University. In the course of their brief conversation, the educator said to the preacher, “Doesn’t it ever bother you that people wonder why you do not have a PhD?” Whereupon the preacher replied, “I’d rather have them wonder why I don’t have a PhD than why I do have a PhD.” Among those present at the seminar that day were our chair, Jack Rigney, and Gertrude Cox, director of the institute of statistics in North Carolina. Neither of these two people had a PhD!

After the seminar, Jack Rigney came up to me and said that he and Gertrude Cox wanted me to stay on at Raleigh to complete the work necessary to earn a PhD, and that the institute would provide the funding.

On January 31, 1978, the Federal Register reported that the U.S. Food and Drug Administration would require a new warning on the patient insert in all packages of oral contraceptives. In an apparent attempt to emphasize its importance, the warning was to appear in bold print at the top of the insert that listed all the contraindications to the use of this medication.

After I analyzed the scientific studies the Federal Register indicated had been used by the FDA in reaching its decision, I concluded that these studies did not support the language in the warning. I presented my case to my superiors and I proposed, further, that our industry call for hearings on the matter before a committee of Congress.

My proposal was outrageous! It is not customary for any industry to request a public hearing before a committee of Congress. On the contrary, most industries would rather appear to be dragged, kicking and screaming, before the U.S. Congress. One of our attorneys challenged me by asking for the strategy I might use to respond to the battery of physicians, obstetricians, and gynecologists the FDA would likely bring to such a hearing as witnesses on their behalf.

I said that I would counter the FDA’s panel of witnesses with two or three witnesses of my own, who were among the most outstanding women statisticians in the United States. By then, I knew my two principal witnesses would be Professor Gertrude M. Cox of Research Triangle Institute in North Carolina and Professor Jean Dickinson Gibbons of the University of Alabama.

The hearings were held in Washington on October 4, 1978, before the Subcommittee of the Committee on Government Operations of the House of Representatives. I was the lead witness.

At the end of my testimony, I presented the written and signed statement by Gertrude Cox. Gertrude Cox did not come to Washington to testify. She could not, because she was seriously ill and in the final stages of leukemia.

Just one week before the hearings, I visited her in her room at Duke Memorial Hospital in Durham, North Carolina. She really wanted to go to Washington to present her own testimony, but her physician would not permit it. She was angry and frustrated, and she wept openly. Then she reread her written statement, signed it, and instructed me to present it on her behalf to the subcommittee. Within three weeks of our visit, at age 78, she succumbed to her disease.

Donald Kennedy, director of the FDA, was the only witness for the FDA. By the time he rose to testify, the subcommittee knew what I had suspected all along: The findings of the FDA statisticians were identical to mine and to those of the other statisticians I had brought as witnesses.

—Marvin Kastenbaum, MS ’50, PhD ’56

“Gertrude’s impact on my professional career and my family life was huge.”

Gertrude’s classes in design of experiments were “alive.” She skillfully guided the dialog to follow the text. One day, we were discussing alternative sampling plans for the study that would follow the current class example. We agreed that using graduate students would not be optimum. However, it would be the cheapest, because the graduate students would not be paid.

Her very few exams were dominantly made up of homework assignments. If one did the homework, it was acceptable to submit the prior work with a few notations.

During the ’53/’54 semesters, I was Gertrude’s data analyst for her outside contracts. The first time I submitted my time, I included only the time at my desk. I was concerned that my “thinking” time was not included. Gertrude was quick to tell me she was having the same problem and asked for my suggestion. I suggested doubling my desk time. After a broad smile, we agreed that would be my rule and hers for the future.

Some of us in the 1954 PhD class were the first NC State statistics graduates to take a job in industry. A few years later, Gertrude organized a two-day symposium on applied statistics and invited all of us to return and talk about our work.

Well into my career, the NC State advanced degree graduates honored Gertrude with a dinner held at an annual ASA meeting. Her gift was the best of the new concept voice recorders. As we assembled, each person spoke a few words into the recorder in a separate room. The gift was presented in a gift box “over stuffed” with paper, well wrapped with an orchid on top. As Gertrude worked her way to the bottom, her many comments were being recorded. When she recognized that the box was empty, she was presented with the recorder and all stood to applaud.

I too was among those who visited Gertrude in her hospital room. Each visit, she promptly showed me her day-to-day plot of her white cell counts. And I was expected to comment. Compliments, yes, but not an easy task.

I want the readers of this tribute who contributed to the Gertrude Cox Fellowship Fund to know that the goal was surpassed. Most important, Gertrude met the first fellow, a woman. Gertrude’s impact on my professional career and my family life was huge. Her picture is on the clipboard I use every day at my desk.

—Mike Free, MS ’52, PhD ’54

“What a wonderful lady!”

My warmest recollection of Miss Cox occurs in 1978, shortly before she passed away. I had acknowledged her on the opening page of Statistics for Experimenters, John Wiley 1978, and she telephoned me from her hospital bed to say thank you. What is most memorable is that I quickly got the impression that Miss Cox wanted to continue to chat. She simply did not want to put down the phone. And so we chatted on, reviewing our varied histories together. What a wonderful lady!

Another strong recollection is being present in that Patterson Hall basement classroom under the steam pipes, watching Miss Cox lecture. There were occasions when she would come to class hurried and unprepared and thus would struggle through the first few minutes. But she always found the way to make the points at issue … you could watch her thinking at the blackboard.

And then there were the graded homework papers, with Sarah Porter and H. P. Andrews as her graders. Importantly, Miss Cox provided me with my very first full-time job. I was hired as an assistant to Bob Monroe and worked in the institute computing lab in the days when obtaining sums of squares required using an IBM punch card sorter and tabulator. Oh, how we celebrated the day the IBM multiplying punch arrived!

And of course, I would not have met George Box as a graduate student if Miss Cox had not invited him to the institute. What a profound influence she has had on my life. She even managed the “cutting the wedding cake” party following my marriage to Tady. Wow!

—Stu Hunter, MS ’49, PhD ’54