Key Takeaways from the Panel on Tenure Processes for Data Science Faculty

24 NOVEMBER 2023

By Stella Min

The ADSA Career Development Network recently hosted a discussion about tips for navigating the tenure process for data science faculty. The panel included Drs. Magda Balanzinksa, Jeffrey Blume, and Rebecca Nugent. Learn more about each panelist and some key takeaways from the career panel below.

👉Don’t forget to check out our previous panels to learn more about career paths in data science.

ADSA's career panel on tenure expectations for data science faculty

About the Panelists

Magdalena Balanzinksa

Dr. Magdalena Balazinska is a Professor and the Bill & Melinda Gates Chair and Director of the Paul G. Allen School of Computer Science & Engineering at the University of Washington (UW). Her research interests are in the field of database management systems. Her current research focuses on data management for data science, big data systems, cloud computing, and image and video analytics. Previously, Dr. Balanzinksa was the Director of the eScience Institute, the Associate Vice Provost for Data Science, and the Director of the Advanced Data Science PhD Option at UW. She is an ACM Fellow and holds a Ph.D. from MIT. She has received several awards, including the inaugural VLDB Women in Database Research Award, an ACM SIGMOD Test-of-Time Award, an NSF CAREER Award, and two Google Research Awards.

Jeffrey Blume

Dr. Jeffrey Blume is a data scientist, administrator, and biostatistician, with vast experience building and leading academic programs. He is the Quantitative Foundation Associate Dean for Academic and Faculty Affairs in Data Science at the University of Virginia (UVA) School of Data Science (SDS). Previously, he served as Director of Graduate Education at Vanderbilt’s Data Science Institute and Vice-Chair of Education in the Department of Biostatistics and was a tenured Professor of Biostatistics, with secondary appointments in Biomedical Informatics and Biochemistry at Vanderbilt. He proposed and established the M.S. in Data Science at Vanderbilt University and established the Master of Public Health program at Brown University. His areas of research and collaboration include radiology, cancer diagnosis and prediction, nephrology, translational biomedicine, fMRI, and women’s health. Dr. Blume holds a Ph.D. in Biostatistics from Johns Hopkins University and a B.A. in Statistics from the State University of New York at Buffalo.

Rebecca Nugent

Dr. Rebecca Nugent is the Stephen E. and Joyce Fienberg Professor of Statistics & Data Science and Head of the Carnegie Mellon Department of Statistics & Data Science.  She received her Ph.D. in Statistics from the University of Washington, her M.S. in Statistics from Stanford University, and her B.A. in Mathematics, Statistics, and Spanish from Rice University.  Dr. Nugent is currently on the leadership team for the NSF AI Institute for Societal Decision Making and has expertise in designing and implementing data science/AI professional development programs for business leaders in industries including health care, finance, automotive/manufacturing, and life sciences. She was the faculty Co-Director of the Moderna AI Academy and the Founding Director of the Statistics & Data Science Corporate Capstone program, an experiential learning initiative that matches groups of faculty and students with data science problems in industry, non-profits, and government organizations.  She has won several national and university teaching awards including the American Statistical Association Waller Award for Innovation in Statistics Education and serves as one of the co-editors of the Springer Texts in Statistics. She recently served as the co-chair for the National Academy of Sciences (NAS) study on Improving Defense Acquisition Workforce Capability in Data Use and served on the NAS study on Envisioning the Data Science Discipline: The Undergraduate Perspective.  Dr. Nugent has worked extensively in clustering and classification methodology with an emphasis on high-dimensional, big data problems and record linkage applications.  Her current research focus is the development and deployment of low-barrier data analysis platforms that allow for adaptive instruction and the study of data science as a science.

Key Takeaways from the Career Panel

The panelists shared several suggestions for prospective and current junior assistant professors who specialize in data science to prepare and navigate the tenure process. Much of the advice emphasized communicating the impact of one’s research and contributions to data science more broadly. Specific strategies for getting buy-in and support from colleagues will depend on the type of position (e.g., joint appointment; teaching track) and the program’s expectations and requirements for tenure and promotion, which vary widely by institution.

More tips for data science faculty

Check out our panel on tips for applying for data science faculty positions to learn more about the structure of data science programs

Understanding the Promotion and Tenure Process

Every college and university has a unique set of expectations for promoting tenure-track faculty. These requirements are often closely tied to the structure of the program. For example, the Paul G. Allen School of Computer Science & Engineering at UW does not offer a formal degree in data science but many specializations within the School highly overlap with data science. Faculty at the intersections of these areas must meet the tenure expectations of their home department if they have a full-time appointment. Faculty with joint appointments must meet requirements that satisfy the expectations of the units that fund their position. Within the Paul G. Allen School of Computer Science & Engineering, faculty members are evaluated on three high-level categories: (1) research and fundraising, (2) teaching and advising, and (3) service. There is an understanding that one’s impact with respect to these three categories may look different depending on the faculty member’s disciplinary background. For example, publications in peer-reviewed journals are not common practice in computer science. If the faculty member under review comes from a discipline that emphasizes peer-reviewed studies, they shouldn’t assume that the tenure committee will know this. The faculty member under review should explain to the committee that peer-reviewed studies are a benchmark of scholarship within their discipline and highlight high impact journals in which they have published their research. This will enable the committee to assess the merits and scholarly impact of their work more fairly.

The tenure process for traditional tenure-track faculty in the Department of Statistics & Data Science at CMU is similar to UW’s School of Computer Science & Engineering. However, the promotion process is slightly different for teaching-track and research-track faculty. These positions are fully funded by the department and the process for promotion is similar to those who are on the traditional tenure track, but evaluations are concentrated on teaching or research, depending on the role and tenure packages incorporate different requirements. For example, for teaching-focused roles, the department solicits internal and external letters of support. Letters of support for traditional tenure-track faculty are external only.

Compared to the two programs described above, the promotion and tenure (P&T) process is structured differently at the University of Virginia (UVA) School of Data Science (SDS), which mostly hires faculty with full-time appointments within SDS. The SDS evaluates a faculty member’s scholarly contributions to the discipline of data science with an understanding that the form of contributions (e.g., publications in scientific journals vs. publishing books; open source software) will highly depend on the disciplinary background of the faculty. The SDS also acknowledges that the form of scholarly work will impact the timeline required to produce it. The tenure committee considers these differences when evaluating the impact of each faculty member who is under review. To understand the significance of a faculty member’s work and to supplement letters of support, the tenure committee within SDS can solicit opinions from others to help situate and bolster one’s scholarly impact on the field, to fairly assess their contributions.

One key aspect of tenure expectations that cut across these different programs was an emphasis on the impact of a faculty member’s work on society and their ability to work on problems that cut across disciplines. The interdisciplinary nature of this work can be difficult to conduct and communicate but is essential for drumming up enthusiasm and support from the scholarly community. The panelists encouraged junior faculty to meet with colleagues as often as possible to share updates about projects and their impact on science, pedagogy, and society. These are the people who will eventually write letters of support, so it is important for them to understand the broader impact of this work in order to convey it others. For those who are or will soon be preparing tenure packages, the panelists recommended dedicating a considerable amount of space within personal statements to communicate the broader impact of your work. Be sure to point to sections within the CV that highlight specific details that support these claims, such as the prestige of a scholarly journal or the number of downloads of open-source software. The main takeaway is to paint a very clear picture of who you are as a scholar and what you are doing to push the boundaries of the discipline in a way that benefits society.

The Tenure Committee is Rooting for You

The panelists acknowledged that P&T processes can place a lot of pressure on faculty members to advocate for their work in a system that hasn’t evolved at the same pace as the discipline itself. Increasingly, leaders of data science programs are working hard to usher in changes that recognize the important work of data science faculty that falls outside of traditional metrics, including the creation of open-source software and reproducible research and developing new curriculum on the foundations of data science. Many senior faculty also support the adoption of these more inclusive metrics and are frequently working with leaders to advocate for them. The panelists emphasized that junior faculty are not alone. There’s a whole team that is working behind the scenes to support them.

If you’re a junior faculty member who feels that this isn’t quite the case in your department or school, especially if you have a joint appointment, Dr. Blume recommended sharing conflicting expectations with both academic units to help align them. For those who are on the market or soon will be, Dr. Nugent recommended paying close attention to how the heads of different units communicate with each other. Poor communication and misalignment of expectations could be a red flag.

Advancing Data Science

Many data science degree programs were launched in response to the industry’s need for skilled data practitioners. Now that the discipline has matured, there’s an increasing focus on how to advance the discipline of data science. The panelists believe that the discipline must move beyond applications of data science methods and begin elucidating how to best train the next generation of data scientists and accelerate innovation. In other words, the discipline must engage in the science of data science. This includes a greater focus on the optimization of data science, as Dr. Nugent phrased it. What are the optimal ways of organizing and collaborating in interdisciplinary teams to solve data science problems? The SDS at UVA is advancing the discipline by carving out a unique niche that is reflected in classes that draw students from across disciplines to learn foundational concepts of data science that are not taught in other domains. Dr. Blume described the vision as one that distinguishes data science from statistics, where many disciplines have created their own statistics modules.

To prepare for this future, early-career scholars must understand where data science is uniquely situated and help develop and teach a curriculum for students with diverse interests and career aspirations. The curriculum must be unique. It should not repeat the same concepts that are taught in other disciplines. This is essential for enabling students to learn foundational concepts faster and apply to work and research, which will lead the way to greater innovation.

The interdisciplinary nature of data science also requires strong communication and presentation skills, in addition to the ability to work effectively in teams. These are skills that every data scientist must cultivate and model for students and junior-level colleagues. Early and frequent exposure to interdisciplinary projects is essential to developing and honing these skills. For those who are early in their graduate programs, Dr. Nugent recommended getting involved in small projects that expose you to interdisciplinary teamwork. However, Dr. Nugent also cautioned against going too deep into one specialty unless you’re confident that it’s where your passions lie. Data science requires a broad knowledge of many different areas. Dedicating a lot of time to one specialty can come at the cost of learning several different skills that would make you a strong data scientist. If you’re further along in your graduate studies, you can still seek out opportunities to gain hands-on experience with projects that tackle data science problems but, at this point, you will need to focus on your dissertation topic. Both Dr. Nugent and Dr. Blume said that while postdoc positions are not essential for data science faculty, they can be a good option for anyone who is switching fields and wants to develop greater expertise in data science.

Summary

The career panel offered several helpful tips for current and prospective faculty who are seeking to understand tenure expectations within data science programs, including how the process can vary depending on the structure of the program and common metrics of evaluation and how they’re changing. The panelists also shared their visions of the future of data science and how early scholars can prepare.

This blog post highlighted some of the key takeaways and insights that were shared during the discussion. Watch the full recording on our YouTube channel to learn more details about the programs that the panelists represent, including teaching loads, credits for developing new courses, and the disciplinary focus of courses taught by faculty with joint appointments.

Looking for More Blog Posts?

Check out our Community Blog.