Paper 4657 -2020

Reaching Your Potential as a Statistician and Leader

Gary R. Sullivan, Espirer Consulting

ABSTRACT
In year's past, a popular question in interviews, mentoring sessions or goal-setting
discussions was "Where do you hope to be in five or ten years?" Many statistical leaders
will tell you that the path of their career was not what they planned, if there even was a
plan. But they can tell you some skills and/or experiences that were instrumental in helping
them achieve advancement and success. So, perhaps a better question is "How will you
advance your skills so that you can grow as a statistician, take on new roles and
challenges, and continue to contribute to your organization and/or profession in a way that
is fulfilling and rewarding?" Simply put, "How will you reach your potential as a statistician
and leader?" This presentation will provide insights and guidance to this question through
personal experiences and leadership study. Concepts that will be discussed include
networking, business acumen, strategic thinking and teamwork. The presentation will also
provide ideas for actions you can take to move forward in reaching your potential as a
statistical leader.

INTRODUCTION
Several years ago, I spoke to a group of graduate students on the topic of leadership study.
These were very astute, enthusiastic learners who lacked professional experience but were
obviously ambitious and driven. Since it is quite a leap for them to go from “student” to
“leader”, I described the career progression in Figure 1 which made sense to them.

Leader

Collaborator

Consultant

Analyst

Figure 1. Career progression for statisticians.

1
I explained the statistical roles as follows:
• Analyst: Practiced in the analysis of data
• Consultant: Skilled in providing guidance upon request for the generation, analysis and
interpretation of data
• Collaborator: Working jointly on an activity or project while also bringing expertise in the
generation, analysis and interpretation of data

I intentionally left out the explanation of leader (for now) to make the following points:
• Most statisticians or data scientists understand the progression and can see the
differences between the roles
• Most have probably gone thru some of this progression in their own careers. (I certainly
did.)
• Finally, reaching the role of “collaborator” in the eyes of your business partners is a
strong accomplishment.

But statisticians should – and many do – strive to be leaders. Several questions

immediately come to mind:
• What is a “statistical leader”? How is leadership defined?
• How do you become a “leader”?
• How do you know if you even have that potential?

I will address all these questions in this paper but will particularly explore the second
question. And let me make clear that “leader” is not a title or something that you are
granted. Rather, it is something that you will work to achieve and to maintain. It is not
“tenured”. I will warn you that the answer to becoming a “leader” is not anywhere near a
well-defined equation or process, though there are specific steps that you can take.

Ultimately, my goal is three-fold:

1. To de-mystify the concept of leadership
2. To accelerate your leadership pursuit
3. To help you reach your potential as a statistician and leader

I will do this by introducing concepts and skills that effective leaders possess, and I will
explain these through observations, examples and experiences. Through these experiences,
I will demonstrate that leadership lessons are on display most every day in your
professional life. The “school” of leadership is in clear view – you just have to become a
student: observe, experience, reflect, and act. I will finish with some guidance on
leadership development and specific actions any statistician can take to reach their potential
as a statistician and leader. Let us start with some basics of leadership and some deeper
thinking about what leadership is.

Before we get into some leadership basics, let me briefly share a bit about my interest and
expertise in this topic. I have been passionate about leadership development for
statisticians for over 10 years. In 2009, I was charged with creating and sustaining a

2
leadership development program for our statistics function at Eli Lilly and Company, which I
did thru 2017. Since that time, I have developed and administered multiple leadership
training courses to hundreds of statisticians and other professionals on the topics discussed
here as well as many others. I retired from Eli Lilly in 2017 and started my own business
focused on leadership development for statisticians.

LEADERSHIP BASICS
What is leadership and what does it mean to be a statistical leader? Establishing an
understanding of leadership is always a good starting place as it can provide direction for
your study and learning, much like a vision statement. Much like an organization vision, a
leadership definition is what you strive to attain in terms of a skill and/or ability. Let’s first
start with the word leader. Most definitions state something similar to “a person who
commands, directs or has authority over a group or team.” These definitions are more
consistent with assigned leaders, which typically indicates a formalized role having a defined
title along with some authority over people, money and/or resources. Examples include
roles like Vice-President of Statistics & Data Science or Director of Statistical Services or
Statistics Department Chair. But these roles are few in most organizations and the
definition of an assigned leader is rather narrow and doesn’t apply to technical statisticians
whose responsibilities are not dominated by administrative tasks. 2012 ASA president, Dr.
Bob Rodriguez, provides a broader, more robust view of leadership in his President’s corner
article from the Amstat News (February 2012). In it, Dr. Rodriguez states:

“Leadership ability is a prerequisite for the growth of our field because statistics is an
interdisciplinary endeavor and our success ultimately depends on getting others to
understand and act on our work.”

Dr. Rodriguez shares a perspective on leadership that speaks more to what we call
emergent leaders. Emergent statistical leaders are those who have no authority or
command over people or resources but can still get their colleagues or teammates to buy
into their thinking (i.e., a statistical approach, an alternative analysis or study design). In
other words, for technical statisticians, leadership is about influencing your business
partners. (See the inset below – “Failure is not an option” – for a more visual reference and
depiction of emergent and assigned leadership.)

I would offer the following three definitions of leadership, from simplest to most
comprehensive:
• The ability to influence without authority
• The ability to get others to understand and act on your ideas
• The ability to consistently deliver value to an organization or cause by inspiring others to
take a specific direction when they truly have the freedom or choice to do otherwise

This last definition is one that was first created by our leadership development team at Eli
Lilly in 2010 that I have updated in recent years. This brings us to an important premise for
this paper: If, through your work over your professional career, you reach a point where
you continuously demonstrate that you meet the conditions of this definition, you are not
only a leader but you are achieving your potential as a statistician and likely pushing it to
greater levels than you thought possible.

3
 Failure is not an option: An illustration of emergent and assigned leadership

It’s often better to show (or see) an example of a concept or skill rather than try to
explain it. A great depiction of both emergent and assigned leadership can be seen
in the 1995 movie “Apollo 13”. The film is based on the true story of the aborted
1970 Apollo 13 lunar mission.

One of the famous scenes is remembered by the line “Failure is not an option.” The
phrase is spoken by Ed Harris, who played Flight Director Gene Kranz in the movie.
In the scene, Harris , assigned leader, challenges his team of engineers by stating
that, under the current conditions, the astronauts will not make it back to earth.
John Aaron, an engineer played by Loren Dean, steps forward and offers a more
pessimistic outcome before proposing a radical solution. Watch as he “emerges”
literally and figuratively as a leader. (This engineer could pass as a statistician.)

For statisticians and quantitative scientists, the scene provides powerful imagery of
what emerging leadership looks like. This type of leadership doesn’t usually play out
in one minute, but the scene does serve as a mini version of what can be required to
influence and lead. Review the scene after reading the paper – it can be found
easily on the internet (LINK) - and you will see several skills on display from both
leaders. Remember that these types of scenes are frequently on display in your
professional and personal experiences. You just need to look out for them.

THE LEADERSHIP JOURNEY BEGINS

Where do you start on your leadership journey? If you are in a professional role with
responsibilities of a statistician or data scientist, you have already started. And the
leadership concepts and skills I will discuss are likely on display all around you. Some of
them you will already be skilled at but most you will have to further develop. Most

4
statisticians learn these skills through their experiences – successes & failures, trials and
tribulations, starts and stops, challenges and achievements. It will take time and there isn’t
a defined end to the journey. But by recognizing what leadership is, committing to its
study, and reflecting on your observations & experiences, you can accelerate your learning
and growth resulting in a more rewarding and impactful career.

I had a long, rewarding 28+ year career at Eli Lilly and Company that included multiple
technical roles and management roles. But my path is not the one I want you to necessary
follow for a few reasons. First, every career is different, and every person needs to find
their own way in their own way. Second, and more important, my recognition of leadership
and its importance didn’t really start until the second half of my career. And my deep
understanding & study of leadership did not occur until the final third of my career at Lilly,
much later than I would recommend. Having said that, my career is also proof that it is
never too late to begin that study. So, if you are past the middle of your career, it’s
certainly okay for your journey to begin now.

What I will do in the sections ahead is to share leadership lessons, skills and concepts
through relatable experiences, even though, at the time of many of these, I didn’t
necessarily equate them to leadership. Rather, I saw it as the “right” way or the “most
effective” way to resolve an issue, solve a problem, or move a group of people forward. Or,
in some cases, the “wrong” way or “most unproductive” way. One of the more important
points I will make is the power of observation and personal reflection in developing
leadership skills. I will also recount for you the experience that truly enlightened me
regarding leadership.

In each section, I will clearly connect the skills & concepts to the definitions of leadership
and will introduce insights that I gained later through personal study that support these
connections and provide questions for your own reflection and study.

COMMUNICATION: THE FOUNDATION FOR LEADERSHIP

I was hired as a statistician at Eli Lilly and Company in 1989. I still remember my interview
with my first supervisor. He was positive, direct, engaging, informative, listened
attentively, asked good questions and gave thoughtful responses to my questions. He was
a big reason I accepted the job. Was he a good interviewer, a skillful communicator or an
effective leader? At the time, I thought he was just a good interviewer. Looking back, he
demonstrated all three during the
interview process. Interviewing is
an exercise in leadership – you are
trying to convince someone to take
a specific direction (i.e., hire into
the company) when they have the
choice to do otherwise. And
communication is the primary skill
involved in the process.

Communication is the foundational

dimension of leadership. The
example of my interview, simple as
it may be, captures many aspects of communication that are necessary for effective
leadership. You will see through all my examples and shared experiences, the importance
of communication and the role it plays in leading. I will make a point to explain the

5
different facets of communication in some of the examples. But let me share some
important aspects of communication here that will be seen throughout the paper.

When people hear “communication”, they immediately think of speaking. Can you explain
your idea? Can you skillfully engage in a debate or discussion on an issue? Can you give a
clear technical presentation or a convincing proposal? Yes, that is a part of communication.
But the other aspects of communication, which are just as critical as speaking, are listening,
writing and actions – small and large. Let me share a question for reflection on each:
• Do you listen in a way to truly understand the intentions, needs and emotions of your
customer, collaborator or business partner?
• Does your writing (and speaking, for that matter) reflect a careful consideration of the
knowledge and position of your audience, and does it strive to appropriately “move”
them to your thinking (i.e., get them to change their thinking or adopt your idea)?
• Are your actions consistent with your spoken or written word and, more importantly, do
they model the leader you are striving to be?

In his book, The Power of Communication, Fred Garcia writes, “The only reason to
engage an audience is to change something, to provoke a reaction. Effective
communication provokes the desired reaction; ineffective communication doesn't.” If you
reflect on the three definitions of leadership I shared earlier, this view of communication is
consistent with all three. This underscores the importance of communication in effective
leadership, including interviews.

Let us move on to some examples and experiences that highlight the other critical
leadership skills and concepts.

THE POWER OF TRUST

My first role at Lilly was as a statistician supporting product development and
manufacturing. This is not the typical role for a statistician in the pharmaceutical industry
as most statisticians support some aspect of clinical trial design and analysis. It was
atypical enough that I was referred to by what I was not, rather than what I was – a
Nonclinical Statistician. In this role I consulted with scientists, engineers and production
specialists to help them optimize processes, resolve manufacturing problems, or
analyze/interpret data to support regulatory filings or submissions. I want to share an
experience that emphasize the importance of communication and trust as the foundation for
influence and leadership.

One of the first significant projects I worked on was to optimize a tablet formulation for a
weight loss drug (which never made it to the market). To understand the critical factors at
a large scale, our formulation team traveled to a manufacturing site where we planned to
execute a Plackett-Burman screening design to understand the effect of various process
factors on tablet properties. I didn’t fully understand the cost of this study at the time –
more on that later – but it was probably close to 0.25M $ when you accounted for materials,
people resources and equipment. The day before we had planned to begin the study, the
process technician, who I had come to know well, asked if we could discuss the study. He
had never applied experimental design in any formulation study and he went on to tell me
that several of the dozen or so experiments wouldn’t work. In other words, they would
produce material that could not be compressed into tablets so we would not be able to
measure chemical and physical properties. In statistical terms, the experiment would have
a significant amount of missing data. In small experimental designs, this creates issues

6
with correlation and variability in model estimates. We proceeded to adjust the factor levels
so that, according to his expertise, we would get results – and we did. The experiment was
a success.

On the surface, this is unlike many interactions that occur between statisticians and their
business partners. But, upon reflection years later, I would offer the following questions:
• What if the technician hadn’t come to me with his concerns?
• Was he sure that I would listen to him? Why did I?
• Was I leading him or was he leading me?

As I mentioned earlier, I came to know the technician well before this critical study took
place. We met for lunch and coffee often and discussed any number of topics ranging from
statistics and formulation to organizational politics and personal principles. We came to
know and trust each other. Would he have come to me regardless? Perhaps. But when
trust is present, we believe that others will act in our best interest or the collective interest
of the group. Our trust in this situation provided mutual benefits. First, the process
technician trusted me to the point that he would use a method he had never used before on
a $250,000 experiment. He did not have the total authority to refuse to, but he probably
could have convinced the others that we shouldn’t. He had 20 years of experience and I
had less than two. And I trusted him to change the levels of some experimental factors on
a study that was reviewed and approved by several of his peers, some more senior than
him.

Trust is an essential component of effective leadership. Trust is business currency. It gains

you access, speeds decisions, creates honest dialogue, affords risk-taking and allows you to
make mistakes. How do you convince someone “to take a specific direction when they have
the freedom or choice to do otherwise”? You get them to trust you first. If you want
someone to “act on your ideas”, start by winning their trust. How do you do that? Through
your actions (i.e., your communication as explained above), you demonstrate the following:
• Competence – are you skilled in your field or the area of mutual interest?
• Care – can you help them achieve their goals first?
• Character – are you principled, ethical and consistent in your approaches & beliefs?

Here are some questions for reflection:

• Have you earned the trust of colleagues and business partners?
• If not, what do you have to do or change to earn that trust?
• What role does each aspect of communication play in building trust?

THE IMPORTANCE OF NETWORKING AND BUSINESS ACUMEN

About five years into my career, I took a role providing statistical support to the Bioprocess
Research & Development organization. This was a group of 150+ scientists and engineers
who were responsible for taking a large molecule (an active biologic) that showed promise
in treating a disease in animals and developing a process to make it in larger quantities to
support human clinical trials and ultimately, if all went well, to make enough formulated
drug to support the market place. The organization included a wide range of people
including molecular biologists, micro-biologists, analytical chemists, bio-chemical engineers,

7
protein chemists, process engineers, and process technicians. My responsibilities were to
help them design studies, analyze and interpret data and provide training as needed.

There were two experiences in this role – both personal “failures” to a different extent - that
helped me move toward being an emergent leader. These highlight the importance of
networking and business acumen in emergent leadership.

When I began this role, I had some contacts and regular customers from my previous role –
a much broader one – and they kept me busy for about a year. As those projects
completed or those customers moved on, I felt that my services were being underutilized.
In any given week, there were dozens of experiments being conducted and hundreds if not
1000’s of data points being generated. Why weren’t people knocking down my door? I
gave presentations on the benefit of statistical methods that were well-received, but that
didn’t change much. Since the projects weren’t coming to me, I realized that I needed to
find them – to network. I needed to go out and talk to individual team leaders, senior
scientists and even managers to understand what they were doing, what problems they
were trying to solve, and convince them to let me try to help. I thought my communication
skills were good – I could answer questions, present
analyses, and teach methods – but this was taking it to
another level, much like being a salesperson. That was
well outside my comfort zone. I was and am an introvert.
I never signed up for having to go out and “sell” my
services and ideas to customers, but that’s what I needed
to do to be successful. I struggled but kept working at it
and eventually got to a point where (1) I was comfortable
managing a conversation with a scientist that could go in
any number of directions, and (2) I was able to get
myself into more team meetings where I could learn the
science and start to influence study designs. Most
importantly, doing this challenged my communication
skills – especially listening and speaking “on the fly” –
and made me a much better communicator. As time
went on, I became involved in most of the major projects
in the organization and worked at one time or another
with the majority of scientists and engineers in the
organization.

The second experience (and failure) started with an idea I had to apply experimental design
concepts broadly within the area to establish robustness and deeper understanding of our
processes before scale-up and transfer to manufacturing. It was a good idea, but I had no
clue how to move it forward. I mistakenly believed that I simply had to share it with my
senior director – the head of BR&D – and, if he thought it was good, he would take it
forward. When I met with him, he had some questions and seemed to support my general
approach. I left the meeting feeling good and waited for next steps on his part – the
formation of a team, communication of the idea, or some call to action. Nothing happened
and I just figured there were other priorities and projects, and this didn’t rise to that level.
Only years later did I realize that it was not my senior director’s responsibility to take the
project forward – it was mine! But I lacked the skills – the leadership skills to be specific –
to do that. I needed to first build broader support for the idea with technical leaders in my
organization. I then needed to network the idea with technical and administrative leaders in
manufacturing. Gaining sponsorship with senior leadership was also necessary as well as
putting together a strategic plan including potential benefits & value, timing, a proof of
concept proposal and a communication package. This would have been an exercise in the
very definition of leadership: Convincing others that this idea was worthwhile and would

8
deliver value, and that they should dedicate their time and resources to it. I did not have
the skills or the capability to take this on, specifically skills in networking, business acumen
and strategic thinking.

As you develop as a statistical leader, you will be challenged to deliver value for your
organization in more impactful ways. Recall the progression of statisticians from analyst to
consultant to collaborator to leader. Making this progression requires a deeper and deeper
understanding of your “business” – and I use that word very broadly. Depending on your
sector this translates to some of the following: science, financials, policies, processes,
operations, regulations, politics, and culture. Some organizations may provide training or
education in these areas but for most statisticians, it is up to you. Even if you work for the
government, academia, or a non-profit, you still need to understand and be able to answer
the following questions:

• Who are your customers and stakeholders? Do you meet with them individually –
separate from busy team meetings - to understand their interests, priorities and
challenges?
• Does your network reach beyond your immediate customers and collaborators? Which
senior leaders or external partners would be beneficial to connect with?
• What are the goals of your organization at multiple levels – group, function, department,
company? How does your work and the product and/or service of your team contribute
to those goals – at all levels?
Answering these questions requires regular networking to gain strong business &
organizational acumen. This provides you a breadth of understanding that allows you to
impact at more than a tactical level. You can determine how your skills and knowledge can
best contribute to solving problems and addressing challenges at more strategic levels of
your organization. Finally, it gives you enough knowledge of the “business” to convince
stakeholders and business partners of your ideas and approaches, and ultimately gain their
support. This is high-level leadership.

LEADING PEOPLE AND UNDERSTANDING POTENTIAL

Most people see assigned leadership as being different from emergent leadership. But an
effective assigned leader uses the same skills as an emergent leader. I would rather
statisticians see these roles as presenting different challenges than some type of
advancement. Taking on an assigned role, which usually includes people responsibilities, is
a way of challenging yourself and can be a good way to understand and grow your
potential. But it should not be seen as a path for everyone. I will share my experience
transitioning to an assigned role but want to talk first about potential.
Early in my career, I had no interest in supervising people and even said as much. But after
10+ years as a technical statistician and individual contributor, I developed an interest in
supervising people and managing an organization. It opened up a new set of challenges for
me and helped me tap into and develop new skills. I also was able to leverage the
experiences I had as a technical statistician and applied them in this role.
I will talk more about that below, but there are two points to emphasize about my decision.
First, I became a supervisor not because I saw it as a stepping stone but because I had an
interest, perhaps a growing passion, in wanting to lead others in a way where they could
depend on me to coach them, challenge them, and help them be successful. When I took
the role, it was no longer about me but about them. Second, I also saw the role as a way of
challenging myself and going beyond the potential I had as a technical statistician. And this
is the important point. There were other ways to take on a new challenge. I could have

9
taken a role as a clinical statistician. I could have moved into a non-statistical function and
used business knowledge I had gained, perhaps as a project manager. I could have moved
into a different non-clinical statistician role to learn a new science or field. The point is that
I took on a role that would challenge me in new ways, and by doing that in opened up new
potential for me – a potential I didn’t see in myself five or ten years earlier.
After almost thirteen years as a technical statistician, I moved into an assigned leadership
role, taking the position of manager for manufacturing statisticians. The lessons I had
learned and shared above as a technical statistician were readily applicable in my assigned
leader role:
• Skillful communication … all aspects – with reports, business partners, and functional
colleagues
• Building trust … with my direct reports
and business partners
• Networking … with business partners,
but also making this an expectation
for my reports
• Business Acumen … understanding
priorities, processes, and challenges of
manufacturing and gaining knowledge
of how we create value for the
company
• Strategic thinking … for projects both
small and large, thoughtfully using the
knowledge and resources at my
disposal to deliver value for manufacturing

Though I had no experience supervising people, I again used my observations and

experiences as a technical statistician as preparation:

• Before becoming an assigned leader, I had six different supervisors. They all provided
examples of good leadership – and some bad - that I was able to learn from. I used this
learning to determine my principles for leading others based on both what I saw as best
practices and my own personality.
• I had collaborated with hundreds of people at varying levels from many different
functions. Developing a rapport with my reports would be similar to collaborating with
my clients – I could leverage critical skills like communication, trust, and reliability.
• As I had substantial experience as a technical statistician, I could coach my group on
techniques and skills to help them grow and have impact.
• Finally - and this is relevant to all technical statisticians – in many ways you are a
service provider. At times, that’s a disadvantage as we want to be seen as equal
partners and collaborators. But we establish trust by serving others. This provides a
good foundation for being an effective supervisor or assigned leader of people.
Whether you follow a technical path, take the administrative path, or move into a new
organization or new company, the leadership skills you learn as a technical statistician will
help you succeed in your next role. And every new role you take will change how you look
at your potential.

10
CREATING OPPORTUNITIES
What can happen when you develop these leadership skills and apply them together? I
share the following example to show how these skills can be collectively used to first identify
opportunities to add value to your organization and then to provide the leadership to move
the opportunity forward and see it to fruition. I share this not to be boastful – I certainly
missed many such opportunities earlier in my career – but to show what impact a
statistician can have by investing in and proactively applying these skills. (In hindsight, I
wish I had gotten to this point in my career faster, and that is one of the reasons for my
commitment to leadership: to accelerate the development of these skills in other
statisticians.)

Even though I was now in an assigned leader role with people responsibilities, I still saw
part of my role as looking for opportunities for my group to deliver value to the
organization. A year into this assigned leader role as manager of manufacturing
statisticians, I had developed a network of business partners who I would meet with
regularly. At one such meeting, I was talking with the head of Parenteral operations, an
organization that did the final product delivery and packaging of large molecule products
(e.g., insulins) in the Indianapolis area. One of the questions I would typically ask in these
meetings was, “What are the biggest challenges you are facing?” (Note: I was careful not
to ask, “How can my statistics group could help your organization?” because I might get a
specific answer based on that individual’s view of our role, which could be narrow or
limited.) His response was fairly immediate: “Deviations are killing us!” I knew a fair bit
about the processes that deviations triggered and gained a more thorough understanding
after networking with some other areas – operations, quality, and regulatory:
• When a measured property on a final product went outside the control limits, it triggered
a deviation. In most all these cases the measured property was within its specifications,
which were wider than the control limits. Hence, the product was acceptable but atypical
relative to previous batches of product.
• For each deviation, a root cause investigation was performed by a technical services
associate and had to be documented when complete.
• Most deviations came back without a specific assignable cause due to a time lag and/or
backlog. In other words, because the deviations were being investigated days or weeks
after they occurred, it was difficult to discern a cause because operators simply couldn’t
remember what happened and records weren’t detailed enough. (More on this below.)
• The intent of recording and investigating deviations was good – find a cause and correct
or adjust course to improve the process & product. But due to the backlog and lack of
specific causes, there was little to no value being gained.
• The control charting was owned by operations, but my group of statisticians calculated
and updated control limits.

After additional thought and discussion, I proposed that we just stop control charting for the
following reasons:
• First, the control charts weren’t being reviewed in real time. The measured properties
were reported and charted several days or more than a week after the process was
completed. This was not unusual as the assays of the product samples took several
days.
• Second, as stated above, because of the delays - in the measured properties, in the
review of the charts, and in the root cause investigations - the entire exercise was not
adding value.

11
• Finally, the exercise was costing money because personnel were spending their time on
pointless root cause investigations while they could have been making improvements to
the products and process in other ways.

The problem was that established procedures mandated that control charting take place.
Why was it mandated that we had to control chart? After some research by my team and I,
we found that the procedure was created in response to an FDA inspection report many
years prior that stated there was no process for investigating atypical results. Since the
inspection report required a countermeasure, the control charting was put in place.

I now had the backstory which allowed us

to move forward on a solution. After
getting approval from the key
stakeholders – operations, quality and
regulatory - we were able to stop the
control charting for 80% of the final
properties. The stakeholders decided to
keep control charting for the most critical
properties as the reduction would
eliminate the backlog and allow real-time
assessment for fewer properties. The
work improved our regulatory standing
and freed up a significant number of
people resources in manufacturing, quality
and statistics, who could then focus on
truly improving processes and products.

I learned one additional leadership lesson in this exercise along the lines of strategic
thinking. One question that was likely not asked before the control charting was instituted:
What is the cost of implementation and the cost of maintaining this new approach/system?
Had this question been asked and carefully researched, I doubt that control charting would
have been fully implemented. This was a lesson I carried forward and would advise that
you do as well. For any proposal, change, or new process, ask questions about
implementation, maintenance, and unintended consequences. This will stop nonvalue-
adding work from starting or assure that the solution is piloted on a smaller scale before
fully implemented.

This is an example of how leadership skills can be used by statisticians to identify value-
adding ideas and then drive them to completion. Here are a few questions to consider or
reflect on:
- Think about a project or initiative in your organization or company, perhaps one you
are a part of. What leadership skills have you or your team members demonstrated
that have been critical to the progress of the project or initiative? How many can
you come up with?
- Do you have ideas and/or possible improvement that could add value to the business
beyond your tactical responsibilities? How could you move those ideas forward?
What would it take from a standpoint of communication, networking and strategic
thinking?

12
THE HIGH-PERFORMING TEAM
One important topic we have not covered is teamwork. Most all statisticians work on teams
– whether project teams, functional groups, or special initiatives. What are the keys to a
functioning and productive team? How do leadership and leadership skills play a role? Let
me share an example to emphasize some key concepts.

Later in my career, I was fortunate enough to be promoted to the role of Senior Director for
Nonclinical Statistics. In this role, I had responsibility for statisticians who worked with
discovery research, product development, and manufacturing. I became part of the Global
Statistical Sciences leadership team that, under the leadership of a new Vice-President,
would be pursuing a strategy based on innovation and leadership. The leadership team was
comprised of eight people, most of whom had worked together in some capacity.

Before the organization became official, the Vice-President – the assigned leader of the
team – held a four-day off-site meeting for strategy development and teambuilding. I was
indifferent to teambuilding at the time, but I saw it as a way to get to know my colleagues
better and that was a plus. The team building part was a bit more intense than I expected
as it involved several exercises and activities that were new and different. For example, in
one exercise we had to meet with each person privately and tell them any issues we had
with them in the past – perhaps they had unintentionally offended, or mislead, or gone back
on their word. If such an event happened, we had to share it with them. In addition, we
also had to share what we admired about them, something they had done which we
appreciated or saw to be especially going above and beyond. I always believed in being
direct and honest with people, so the exercise wasn’t difficult for me. As the new
organization became official, we continued to have periodic off-sites which included different
team-building events. Only later did I realize the purpose, and much of it had to do with
building trust within the team. Although our Vice-President didn’t really say this, she very
well could have as it captured the essence of her approach to building this team:

“We are pursuing a very aggressive strategy, unlike anything the statistics group has
done in the past. It will require your commitment to this team and to one another. As
we proceed, we will have many important discussions and will need to make the best
decisions on a variety of issues. In order to do that, I will need you to trust each other,
to be honest, and not hesitate to share your perspective on issues and situations. We
must be comfortable having difficult discussions. This will help us make the best
decisions and determine the best path for moving forward on any issue. Even if we
don’t all agree, once we have decided on a path forward, we will collectively support it.
We will not leave the room and second guess. We will hold each other accountable to
the actions and goals we set. If we can do these things, we will be successful as a team
and achieve great things.”

About a year after this team formed, I had an issue with one of my colleagues on the team.
He had acted in a manner that I felt was undermining my organization. I went to my Vice-
President and shared the situation. She didn’t take sides and asked a few questions before
saying what I knew she was going to say: “Now go resolve it with him.” I expected nothing
less, and we resolved it. This team was the most productive and rewarding team I was ever
a part of.

Leading naturally involves moving people toward a goal or inspiring them to support a cause
or initiative. When assigned and emergent leaders can establish a trusting environment,
their team has productive discussions which result in better ideas and decisions. This

13
reinforces commitment, creates accountability and delivers value to the organization. (See
the inset below for a reference for a more in-depth understanding of teamwork.) Without
trust, discussions can be superficial, decisions are sub-optimal or second guessed, team
commitment suffers, and goals are not achieved.

In his book, The Five Dysfunctions of a Team, Patrick Lencioni provides a simple,
powerful framework for building a high-functioning team. Like most of his leadership
books, he uses a fictional story to show the concepts being applied in a believable
situation and then defines and summarizes them in an understandable way.

In my reflections on teams, I have found that the degree to which teams are
unproductive and ineffective is directly tied to number and level of the five
disfunctions. I also agree that establishing trust between team members is a
necessary first step toward strong team performance. The book is an easy read and
is always one of the first books I recommend to aspiring leaders.

The role of communication in effective teamwork is also clear, especially the aspects of
listening, speaking and actions. Building and maintaining trust depends on effective
communication. The success of every subsequent team meeting depended on every facet
of communication. And every communication either strengthened or weakened the trust.
Additionally, the ability of the assigned leader to listen, speak and act was paramount. She
had to set the tone, provide the example, and act in a way that reinforced her plan and
expectations.

Here are some questions for reflection:

• What are the best and worst teams you have ever been a part of? What made them
good or bad?
• Are your current teams productive? What is the level of trust on those teams? Are
there opportunities to build greater trust?
• What role can you play in doing that?

LEADERSHIP TRAINING
What role does formal leadership training play in a statistician’s development? Let us first
consider how a statistician develops technical expertise in a specific methodology (e.g.,
Bayesian methodology). What ultimately makes a statistician a Bayesian expert is a
combination of the following:
• Formal graduate training
• Coaching
• Application
• Professional training (e.g., short courses)
• Personal study (literature, books, articles, presentations)
• Peer discussion
• Applied research

14
There may be other activities not captured here. To develop deep expertise in a statistical
subject, a statistician needs to commit to personal study and learning over a long period of
time. Here is a similar list of activities for leadership development:
• Professional leadership training
• Observation
• Experience
• Reflection
• Mentoring
• Personal study (books, articles, presentations)
• Peer/Group discussion

Training is always a good place to start.

Courses and programs that I have taken,
developed and instructed focus on
foundational concepts, specific skills,
exercises, personal reflection, discussion and
action planning. Training can provide a good
overview and a direction for initial study no
matter where you are in your career.
I took my first leadership training twenty years into my professional career. It was a
corporate course which included professionals from across the company and was a week-
long off-site program. The program focused on a few specific topics but also included case-
studies from both company leaders and community leaders. The program also included
team exercises, plenty of table discussions and individual assignments. In addition to the
learning on the topics and the case studies, I came away with three important lessons from
the training:
• Effective leadership is critical to the success of any organization or cause.
Organizations, no matter how small or large, ultimately thrive or struggle depending on
the quality of their leadership.
• Time spent reflecting on observations and experiences is a crucial component of
leadership development. To truly learn from experience requires deep reflection to
understand the concepts and skills in play, followed by some action or effort to put the
learning into practice.
• The best leaders are life-long learners and true “students of leadership”. No single
course, experience or role can make you a complete leader. Rather, a leadership
endeavor requires a willingness and desire to be a continuous, lifetime learner.
After taking this course, I began to reflect on my past experiences – some I’ve shared here
– and understood the leadership lessons they provided. From that point forward, I viewed
most every experience thru a leadership lens, always looking for a leadership lesson, a
helpful technique, or an idea as to how to be better. Whether a team meeting, a crisis
situation, or an interaction between colleagues, I could usually find some leadership
learning related to any number of the concepts that I discussed in this paper.

15
CONCLUSION
Everyone has the potential to be a better leader. Only by making the effort to study and
improve will you know how much potential you have. I have discussed many of the
necessary skills statisticians must acquire to advance their leadership abilities and
emphasized the benefits of experiential learning and reflection. I offer the following
guidance on how any statistician can realize their potential as a leader:
• Identify the 2-3 skills required to be successful in your current and future role
• Commit to improving at least one leadership skill for six to twelve months
• Find one or two mentors who can help you by assessing your skills and suggesting
actions you can take to improve those skills.
• Look for training as a way to gain deeper leadership understanding and improve specific
skills.
• Form a small leadership discussion group to share leadership experiences and benefit
from the learning & perspective of others.
• Practice, practice, practice – take action to apply what you learn and then reflect on the
experience to continuously improve.
• Start now! It’s never too early or too late.
Vince Lombardi, a famous American football coach is quoted as saying, “Leaders aren’t
born, they are made. They are made by hard effort, which is the price which all of us must
pay to achieve any goal which is worthwhile.”
I agree with Coach Lombardi and will close with the following points:
• Some people are born with more leadership potential than others, but to realize any
potential requires effort. I’ve seen “lower” potential people become better leaders than
“high” potential people by working harder at developing their leadership.
• To re-iterate, leadership for statisticians is a goal worth achieving as it will provide a
rewarding professional career filled with challenges and achievement.
• Achieving leadership requires a commitment of personal study and lifelong learning.
Good luck on your leadership journey!

REFERENCES
Rodriguez, Robert. February, 2012. “Statistical Leadership: Preparing Our Future Leaders.”
AmStat News, Issue #416: pp 3-4.
Apollo 13. Directed by Ron Howard. Universal Pictures, 1995.
Garcia, Helio Fred. 2012. The Power of Communication. Upper Saddle River, NJ: Pearson
Education, Inc.
Lencioni, Patrick. 2002. The Five Dysfunctions of a Team. San Francisco, CA: Jossy-Bass.

ACKNOWLEDGMENTS
Thanks to Alexander Schacht for his review of my paper and his business partnership.
Thanks also to Bill Coar for nominating me to be an invited speaker at the 2020 SAS Global
Forum.

16
CONTACT INFORMATION
Your comments and questions are valued and encouraged. I am also available for training,
mentoring and advising on leadership and leadership programs. My contact information is
given below. You can also connect with me on LinkedIn.
Gary R. Sullivan
Espirer Consulting LLC
317-363-1055
[email protected]
www.EspirerConsulting.com

17