Collaboration with Doctoral Programs

Differences Between Doctoral and Master’s-Level Education

Any discussion of why or how to work with doctoral students probably needs to begin with an understanding of the differences between master’s-level education and doctoral-level education.

Simply put, in a master's program, students learn how to "do stuff", while in a doctoral program students learn how to develop "new stuff to do".

There are broadly two types of doctoral programs that you will likely encounter - the Professional Doctorate (e.g., Doctorate of Business Administration or DBA, EngD or Doctorate of Engineering) and the Doctorate of Philosophy (PhD). While universities offer other kinds of doctorates (e.g., the Educational Doctorate or EdD, Juris Doctorate or JD), these degrees are less likely to be aligned with the analytical needs of your organization¹. In Chapter 2, Table 2.1 provides the classifications of universities and colleges in the United States. Almost all PhD programs will be offered through universities classified as “High” or “Very High” Research Activity. Doctorates are primarily offered by “High”, “Very High”, or “Professional Doctorate” universities (see Table 2.1). Both Professional Doctorates and the PhD

Table 5.1 Distinctions Between the DBA and the PhD Degrees

are considered research degrees, both qualify the graduate to teach at the university level, and both require a dissertation. However, there are important distinctions. For example, consider the distinctions between a DBA and a PhD presented in Table 5-1.

Most master’s degrees require 1—2 years of study, while most doctoral programs require 4+ years of study. However, the difference is not just additional years of study - doctoral students must engage in independent, scholarly research that will make a meaningful contribution to their discipline.

Tlie requirement that doctoral students, both PhDs and Professional, must engage in independent research, where master’s-level students typically do not (although some do), cannot be overstated. At our university, we see student applicants as well as corporate partners who do not appreciate the difference between the doctoral requirement to engage in “research” versus the master’s-level requirement to engage in “projects” - master’s students are more typically “doers”, while doctoral students are (should be) “independent researchers”. In fact, one of the interview questions that we pose for our doctoral program is “What do you see as the main differences between a master’s-level degree and a doctoral degree!'’ Students who cannot articulate a difference are not offered a second-round interview slot.

I tell many of our applicants that in many ways a doctoral student in data science has more in common with a doctoral student in sociology than with a master’s-level student in data science.

The point is that there are common skills doctoral students develop² which most master’s-level students do not (e.g., conducting a literature review, formulating a novel research question, developing a proposal, producing peer-reviewed scholarship, and - the ultimate doctoral experience - defending a dissertation). These experiences and developed skills are consistent to all doctoral programs from chemistry to theology, mathematics to business.

Between the two of us, we have almost 50 years of experience in academia (Bob is older) - where many of those decades have included connecting the classroom experience with organizations seeking analytical talent. Over that time, we have worked with hundreds of analytical hiring managers — and almost all of them ask about some combination of five critical skills. While these skills can be found at the undergraduate and master’s level, they are actually foundational to completing a PhD:

• 1. Ability to innovate and think creatively. At the core of any doctoral program - professional or PhD — is the ability to identify and articulate a novel, non-trivial problem, synthesize what is already known across a wide range of outlets, and develop a theoretical and/or applied framework to address the problem that will inform how people think about the same problem differently in the future. The ability to engage in innovation, creativity, and original research is increasingly a set of skills heavily sought by managers seeking analytical talent across all sectors of the economy.
• 2. Communicating orally, visually, and in writing. We frequently tell our PhD students (actually all of our students) that it does not matter how smart they are if they cannot explain what problem they are solving, how they are solving it, or why people should care. Over the four plus years of their studies, PhD students are required to write research summaries, conference abstracts, journal manuscripts, and, of course a dissertation. Students become “fluent” in writing in multiple formats (it seems like every conference and every journal has their own stylistic preferences) as well as Power Point. As they share their research, PhD students in data science must learn how to present to both technical and non-technical audiences; they effectively have to become “analytical translators” with the ability to communicate with highly technical, computational scientists as well as with less technical, but highly influential decision makers who demand that the results are tied back to the original business problem (i.e., “so what”?). In addition, most data science PhD students will have teaching responsibilities, which is a particularly robust training ground for honing public speaking skills - many of the companies with which we regularly engage specifically ask for resumes for graduate students (master’s or PhD) who have teaching experience because they tend to be stronger communicators and are more comfortable speaking to clients and senior decision makers.
• 3- Leadership. PhD students develop varying degrees of leadership skills through their research and through their teaching. As a teacher, they develop the skills necessary to motivate younger students (typically undergraduates). Through grading and assessment, they develop skills related to evaluating performance and giving constructive feedback. In the context of their research, many doctoral students will work within a larger research lab with faculty and with other students. As they progress through their program, doctoral students are expected to assume greater responsibility in these labs to mentor and support more junior research students. They are expected to take ownership of the knowledge generation process and command authority in their research area.
• 4. Project (Time) Management. Large research initiatives like dissertations require students learn the basics of project management. In the early days of a doctoral program, four years seems like an eternity. However, students quickly realize that with teaching, research, writing, and demanding faculty, four years go by very fast. Pursuing a doctorate is an exercise in project management with students recognizing the importance of developing realistic timelines, milestones, and managing the expectations of stakeholders.

For some reason, doctoral programs seem to facilitate fertility; in our PhD program, almost every married student had a baby during their program. As a rule, this event is typically not integrated into a doctoral timeline and is not part of the expected deliverables - research or otherwise.

5. Collaboration. While some doctoral programs may expect students to work completely independently, interdisciplinary collaboration is increasingly becoming more common; most students work with multiple faculty members, other students, and research sponsors on concept development, experiments, simulations, papers, and presentations. This is particularly true in analytics and data science where students are working across multiple disciplines like computer science, mathematics, business, healthcare, and statistics to solve problems and engage in research teams. As the data science community moves away from the concept of “unicorns” - single individuals who are trained in every aspect of data science - to teams of people who specialize in the different roles across the analytical continuum, the need for collaborative skills and the ability to work productively in an interdisciplinary team is frequently flagged as a “critical” skill.

Figure 5.1 Areas of study for practicing data scientists.

In a 2018 study³, practicing data scientists were asked what they studied in school. The results in Figure 5-1 indicate the relevance of interdisciplinarity and collaboration.

It is important to understand too that the term “data scientist” increasingly encompasses a whole category of positions that require varying degrees of specialization: data analyst, data architect, data engineer, business analyst, marking analyst, and business intelligence expert. In the context of larger analytical projects, all of these roles will be required - and required to work together.