Kelly Zou: Mathematics, Statistics, Data Science, and Dreams
This is the second interview in the leadership series launched by the Biopharmaceutical Section. The first interview was with Gregory Enas. The series includes interviews with statistical leaders among the biopharmaceutical community.
Kelly H. Zou, a PStat® and ASA Fellow, is senior director and analytic science lead of real-world data and analytics at Pfizer Inc. She earned both her MA and PhD degrees in statistics from the University of Rochester and completed her postdoctoral fellowship training at Harvard Medical School. Her undergraduate major was mathematics, along with a minor in physics. Previously, she was associate professor at Harvard Medical School, associate director at Barclays Capital, and senior director and statistics lead at Pfizer Inc. She has authored more than 130 articles and several statistical monographs.
Kelly has been on several professional journal and magazine editorial boards and is the incoming chair-elect and past secretary of the Health Policy Statistics Section; chair and past vice chair of the Statistical Partnerships among Academe, Industry, and Government Committee; and chair of the Committee of Presidents of Statistical Societies Awards Committee.
What is your particular leadership style?
I entered an election for the first time when I was in High School Affiliated to Fudan University in Shanghai, China. Our homeroom teacher and a wonderful mentor, Mr. Dehong Wang, encouraged the students to volunteer by taking charge of student activities. There were multiple roles to be filled, and I entered the race to become a league representative. I discussed with my boarding-school roommates, wrote a speech with solid suggestions about how to improve our organization and connectivity among students, and gave a speech in front of all my classmates. Then the ballots poured in …
Probability became an extremely useful tool that my classmates can still recall tallying up and counting the ballots one-by-one on a blackboard. It was a gut-wrenching moment between the two fierce candidates. After being declared the winner, I helped other class officers on events such as celebrating the New Year, other national holidays, and traditional cultural events such as the Moon Festival.
Fast-forward to 2016. My most recent election was for the incoming chair-elect position of the Health Policy Statistics Section of the ASA. I am very grateful that our ASA members have put their faith in my hands by casting their votes of confidence. Several goals have entered my plan for our section members, which are how to encourage professional statistical activities, improve networking opportunities, enhance communication skills, and provide influences in the policy arena using sound statistical methodology and applications in health care and related fields.
I would like to bring up a useful article that has echoed well in my experience on the topic of statistical leadership. It summarizes the following nine important traits that define leadership:
In my view, an effective leader exhibits a combination of all these qualities. Although these traits would embody a universally admired figurehead, you get the general picture that this leader is an inspiring visionary who is also compassionate and responsible toward the betterment of our society.
I wish for there to be visionaries with determination, innovation, and compliance in the statistical profession and quantitative analytics community, in general. My favorite author, Lu Xun (1881–1936), once wrote, “Hope can be neither affirmed nor denied. Hope is like a path in the countryside: originally there was no path—yet, as people are walking all the time in the same spot, a way appears.”
This summer, the biopharmaceutical blog will feature several prominent statisticians, including Erik Pulkstenis and Christy Chuang-Stein.
A key obstacle for those who are technically focused is that excellent workers may not easily evolve into visionary leaders. Many of us may wonder why it is not so easy to break and then surpass this artificial ceiling by thinking strategically. For example, it can be a dilemma how to juggle and balance multiple tasks in work and in life. The personalities on a team may be varied, and a leader must understand the purpose of a complex task, the main paths taken to achieve the success, and to understand the needs and limitations of the team members.
In turn, a great leader can function as a shepherd, who guides passengers to form a critical mass. Faced with employees with diverse goals, interests, skill sets, and backgrounds, an effective leader takes charge, but collaborates effectively. In my company, for example, the “own it” mentality is highly encouraged for all leaders and employees alike.
I have been fortunate enough to thrive along two quite different career tracks. For about 10 years, I progressed from postdoctoral fellow to associate professor at Harvard. For nearly just as long, I have been working in the industry as first associate director at Barclays and then senior director at Pfizer. Although these are well-known academic and private institutions, my experience has been with these selective few. While at Harvard, I was a member of the faculty taskforce the Joint Committee on the Status of Women (JCSW). The JCSW discussed extensively, particularly in terms of the glass ceiling for female employees and faculty.
I tend to be an extrovert in terms of my personality, except when I am reading a protocol, designing a statistical analysis plan, or writing a manuscript. I generally solicit opinions well in advance, but I also give colleagues or committee members time to discuss on a smaller scale within sub-teams before finalizing the overall task or project.
Since I have been exposed to multinational and multifunctional settings, I also see things from different angles and cultural perspectives. Everyone has a different and innate purpose and mission in life, and everyone must come up with and come to terms with his or her own definition of happiness. As long as life is fulfilling, it will make one’s life content.
On Aspirations and Goals
What are your views on the pursuit of the American dream?
There may not be a universally accepted magical equation that will solve all career-related puzzles, dilemmas, and challenges. In fact, there is hardly a common panacea to guarantee the attainment of happiness if everyone has his or her own version or definition of happiness. A relatively young Steve Jobs once famously asked the vice president at Pepsi-Cola, John Sculley, “Do you want to sell sugared water for the rest of your life? Or do you want to come with me and change the world?” For some diligent workers, selling sugar water is precisely the heart of their pride and joy, and there is nothing wrong with that, either.
As an extrovert and people-oriented person, I get integrated into the environment and culture fairly quickly and seamlessly. I can easily engage in a meaningful chat with new acquaintances through common aspirations and interests. In America, which is a multicultural melting pot, I have not only showed acceptance and gratitude, but also respect and curiosity. I am also realistic and tend to look forward.
I like to set reachable goals and identify the best quality of each collaborator. For example, when planning for a major project, a grant application, or a manuscript, I pay attention to the guidelines, instructions, authorship arrangement, and deadlines. Thus, others will know what to expect of the entire team well in advance. Then, I hope to motivate and spark the creativity in each of my coauthors or teammates, so that the project glues the entire team together. Efficiency is as important as perfection, and I always strive for utmost excellence and efficiency.
Over the years, my coauthors and I have been grateful for the various opportunities that we have had by winning multiple poster awards from the Biopharmaceutical Section and the Scientific and Public Affairs Advisory Committee of the ASA. The coauthors not only developed and published new methodology, but also learned how to present complex statistical work and analyses in visual and succinct ways. After all, solving new problems is a fun way of extending a perpetual learner and scholar to a technically savvy statistician or data scientist in the biomedical field.
However, having aspirations by pursuing the “American dream” may mean different things for different people. Some of us may aim to be leaders, some may aspire to be excellent applied statisticians and data scientists, and some may find theoretical work highly rewarding.
In the Asian culture in which I grew up, for example, we wish for our family members to be completely free of any major illness or catastrophe. To achieve such a balance, we may choose to maintain “the doctrine of the mean” and value harmony, rather than constantly riding the ebb and flow of unexpected tides. Along one’s career journey, a bamboo ceiling may be encountered.
To be successful, one must first become an expert in a particular subject matter or a specific topic area, without being easily replaceable due to the lack of knowledge. Another suggestion is to acquire the useful skill of delegating tasks and not being hands-on all of the time. If an assignment is not within one’s comfort zone, then it is imperative to examine whether the task adds value and, if so, learn more about it and grow from the learning experience.
In reality, however, implicit biases may exist, but must be eliminated. For example, there are the so-called glass ceiling and bamboo ceiling. According to the U.S. Department of Labor, a glass ceiling is “the unseen, yet unbreachable, barrier that keeps minorities and women from rising to the upper rungs of the corporate ladder, regardless of their qualifications or achievements.” The same report states, “Breaking the glass ceiling is an economic priority that this nation can no longer afford to ignore. It is an economic imperative driven by recent dramatic shifts in three areas that are fundamental to business success: (1) changes in the demographics of the labor force; (2) changes in the demographics of national consumer markets; and (3) the rapid globalization of the marketplace.”
What do you think about the idea that a good leader is, first and foremost, a good mentor? When is it appropriate to take a mentorship mentality with a leadership role, and when not?
The key characteristics of an effective mentor are listening and problem solving skills. Nowadays, soft skills can enhance technical prowess. These are all crucial in business acumen and in leadership qualities in the field of statistics, data science, and quantitative analytics.
An important component of mentorship is that mentees should think of the goals they hope to achieve and how mentors could help… It may be useful to jot down a short list of topics on which the mentees may have to seek mentors’ input.
The past chair of the ASA SPAIG Committee and incoming ASA president, Dr. Barry D. Nussbaum, has encouraged our committee to foster active networking and partners across different career sectors. He spoke highly of the impact of mentorship among seasoned and upcoming ASA members. One of the most endearing mentors to many of us was Professor Ingram Olkin.
Well, mentorship has influenced my career path in a profound way. One day when I was about 21 years old, a mathematics professor made a comment, “Kelly, you seem to enjoy probability so much. Have you thought of studying statistics?” Until that moment, I had never thought of statistics outside the framework of mathematical statistics. I was intrigued and excited. At that point, an entire world of probability density functions unfolded themselves in front of my eyes and mind. I dived into the library to devour professors Norman L. Johnson, Samuel Kotz, and N. Balakrishnan’s continuous distributions. They tied so well into my undergraduate mathematics research on Fourier series and differential equations. If I had not met my mentor, I would certainly not have had the exposure to the beauty of statistics in the realm of mathematics, other than regarding it as one of the courses I took.
For those of you who feel the same passion toward mathematical statistics, please read professor Edward Frenkel’s book, Love and Math: The Heart of Hidden Reality, which won the 2015 Euler Prize from the Mathematical Association of America.
Last year, to help with young statisticians, I signed up as a mentor during the last Joint Statistical Meetings (JSM), met with my designated JSM mentee on the night of the Opening Mixer, and attended the JSM Mentoring Workshop. There was great rapport between us as mentor and mentee, respectively, and we discussed her career goals and current statistical pursuits throughout the mixer. By being a mentor outside the immediate scope of work-related projects, it is more rewarding to listen attentively and provide useful suggestions to the mentee.
My company constantly encourages talent development through mentorship paring. In fact, there is a mentorship-matching process in place within our company, which may be utilized by any colleague. I have been quite active in participating in the mentorship process, organized by the Global Asian Alliance (GAA) within my company. A thoughtful and caring GAA mentor from another part of the company became a guiding light. Subsequently, through a friendly “on-boarding buddy” process for a newly hired data scientist, I became a mentor through the formal GAA-sponsored mentor-mentee pairing process and training. I was quite honored to serve as a GAA mentorship panelist.
Generally, mentors and mentees are encouraged to meet periodically, such as once a month. An important component of mentorship is that mentees should think of the goals they hope to achieve and how mentors could help accordingly. It may be useful to jot down a short list of topics on which the mentees may have to seek mentors’ input. In our company, besides technical and applied skills, individual development plans become a blueprint for short-term and fairly long-term aspirations. It may be beneficial to share in confidence some of these goals with the mentor. Furthermore, mentees broaden their horizons by seeking out different mentors with skill sets, technical and interpersonal, that the mentees may discuss and draw inspiration from.
On Life Outside of Work
What do you do with your free time? What are your hobbies? Do any of your hobbies or activities give you a unique perspective for how you conduct yourself at work?
While I was a high-school student, I was a television show producer and a director of the “You and I High School Students” series, which was developed by and broadcasted from the Shanghai TV Station. In addition, I am an oil-painting enthusiast, particularly impressionistic artwork.
Of all interests outside work, I can say that I am a classical music enthusiast. My most favorite composer is Giacomo Puccini, and my most admired tenor, now a baritone, is Plácido Domingo.
I am an avid movie buff ever since I was invited to be on a movie critiquing team during my junior-high-school years, and I especially love watching cartoon movies and TV series via computer animation. If I hadn’t studied math as a major and physics as a minor before pursuing my doctoral education in statistics, I would probably have gone into the multimedia, broadcasting, film, theater, or television business. From time to time, quoting humorously and metaphorically from movie scenes or stage plays becomes a staple of my conversation style.
I recall that my language teacher asked me to join the movie critiquing team when I was a junior-high-school student, and I wrote two pieces a week after watching two movies. I was a playwright who directed several theater plays on stage, as well as comedies.
Of course, there is always a “math geek” lurking. Tracing back historically and through my own PhD adviser and mentor, professor W. Jackson Hall, I have literally become one of the academic descendants or offspring of one of the greatest German mathematicians of the 19th century, Carl Friedrich Gauss (1777–1855).
For example, music notes make me think of white clouds floating in the air, full of Greek symbols. When music and statistics intersect, the creativity is enjoyable! I admire the professor band “The Imposteriors.”
My lifelong goals as a mathematics-statistics-data science whiz exist across space and time, transcending boundaries and limits. Beyond statistics, I have always hoped for many layers, facets, and dimensions within one’s lifetime. Like Brownian motion, little unexpected events would bring out so many facets and colors. One life full of adventures can still be sufficiently satisfying.
In closing, when I have a little free time away from data and deadlines, I am still a dreamer …