Panel Discussion at CSE21 Offers Advice to Mid-Career Mathematicians

By Lina Sorg

The mid-career point is an exciting time for researchers in science, technology, engineering, and mathematics (STEM). As experienced scientists settle into their professions, they begin to think more diligently about future research or occupational directions, professional and personal opportunities, and broader community engagement. A panel discussion at the 2021 SIAM Conference on Computational Science and Engineering (CSE21), which took place virtually earlier this year, offered guidance to applied mathematicians and computational scientists at the mid-career stage. Hans De Sterck (University of Waterloo), Katherine J. Evans (Oak Ridge National Laboratory), Sarah Knepper (Intel Corporation), Damian W. Rouson (Sourcery Institute and Sustainable Horizons Institute), and Mayya Tokman (University of California, Merced) comprised the panel, which attracted over 120 attendees in various phases of their careers.

The session began with a conversation about the seemingly universal fear of plateauing at the mid-career point. Tokman acknowledged that this type of stagnation is more likely to occur for individuals in academia who have had particularly strong early careers and are exhausted from working towards tenure. Upon achieving tenure, mentorship and other support systems within universities are typically less strong than in early-career settings. These factors—when combined with the temptation to relax a bit—can collectively cause burnout.

To combat a prospective slump, Tokman advised attendees to consider their passions and brainstorm ways to best exercise the newfound flexibility of tenure. “Think about possible new directions that might be a bit more risky,” she said. “Being able to say no [is important] as well, because the amount of service requests is going to grow quite significantly.” Increased participation in committees, groups, and other volunteer activities is typically expected of professors with tenure. However, one should also maintain personal research projects, as these endeavors determine promotion eligibility.

Evans affirmed that nearly everybody plateaus at some point in their career, simply due to the evolutionary nature of STEM. “We’re in computing and math, and architectures and other things will change,” she said. “You’ll find that what you’ve been working on will change too. To stay current and fresh, you need to learn more skills.” Acquiring these skills sometimes involves working on larger, more complex projects. Knepper urged mid-career scientists not to get discouraged if they feel like they are accelerating at a slower pace, which often happens with long-term assignments.

A panel at the 2021 SIAM Conference on Computational Science and Engineering (CSE21), which took place virtually in March, addressed the mid-career stage in applied mathematics and computer science. Top row, left to right: Hans De Sterck (University of Waterloo), Katherine J. Evans (Oak Ridge National Laboratory), and Sarah Knepper (Intel Corporation). Bottom row, left to right: Damian W. Rouson (Sourcery Institute and Sustainable Horizons Institute) and Mayya Tokman (University of California, Merced).

Rouson offered a counterpoint to these perspectives and suggested that plateauing does not necessarily have to be a bad thing if it helps individuals slow down and reevaluate their priorities, especially in industry. In this sense, the mid-career period can be an appropriate time to set future goals. “One important thing to consider is, what are you passionate about?” Rouson said. “Does that necessarily align with rising in an organization?” For instance, moving into an administrative position is likely not in the best interest of someone who wants to conduct technical research.

Knepper, however, recently transitioned from software engineering to a managerial role and is enjoying the different responsibilities. Though she spends less time coding, she can now explore strategies to promote productivity, team culture, and diversity and inclusion. This career shift offered Knepper more opportunities to support the types of changes that she wants to see in the workforce.

Discussion then turned to tactics for handling regrets about previous work-related choices. Rather than regret any of his decisions, Rouson chooses to see every problem as an opportunity. Knepper agreed, encouraged attendees to think in terms of lessons learned, and spoke about a challenging project in her past as an example. Although she spent more time on the project than she would have liked before ultimately giving it to another team with a better fit, it taught her valuable lessons about logically managing transitions.

Evans reminded participants to take risks despite the possibility of failure. She also warned that stakes can feel higher at the mid-career stage. “They forget to tell you that taking risks by nature means that some of them will not pan out and you will fail,” Evans said. “When you fail at something, that’s part of the deal.” Coming to terms with the reality that failure is always a potential outcome made it easier for her to process and accept disappointments.

De Sterck admitted that envisioning the effective coexistence between a career and family is sometimes difficult — a concern that is relevant at all levels of experience. Mid-career scientists should carefully consider the ramifications of major workplace decisions to ensure that they align with the expectations of one’s personal life. For example, De Sterck changed institutions—effectively moving his family to a new continent—six months before his second child was born. “In hindsight, maybe that was not the best time to do that,” he said. “It’s not that I think we shouldn’t have done it, but it’s good to be aware of what you’re going to have to go through if you’re in such a situation.”

Increasing demands on one’s time—in terms of both occupation and personal/familial responsibilities—necessitate effective work strategies. Evans advocated for time blocking as the best way to get things done. She suggested that researchers cluster their meetings and schedule several lengthy blocks of time for uninterrupted work; this approach is especially valuable for coding projects, during which distractions like phone calls or emails are especially disruptive. 

As individuals advance in their careers, they may find that balancing “fun” aspects of research with the need to manage and mentor students can become overwhelming. De Sterck recommended that professors always maintain a good sense of their students’ basic technical abilities, and noted that he generally tries to preserve at least one project for which he does a significant amount of the technical work himself. Version control systems like GitHub allow professors and mentors to collaborate and stay connected with students and postdocs, especially as new hardware arises. 

In some ways, having a cohort of multiple students allows more senior mathematicians to partake in a wider variety of projects than at the early-career stage. Yet as researchers accumulate more students over the years, they typically have less time to devote to each one individually. Tokman initially had seemingly unlimited hours to spend with her first graduate student; now she has much less time to split between several graduate students, which was a challenging adjustment. She encouraged attendees to enlist a second advisor or even a postdoc when they have limited time to dedicate to their cohort. “We need to take the idea of raising the next generation very seriously,” she said. “Mentorship in our field is very important.”

Next, panelists discussed the viability of changing institutions at the mid-career stage. Tokman admitted that it is often quite difficult to move within academia at this point, and advised participants to do so before reaching tenure. Academics should ideally start thinking about this type of transition in advance and try to give talks or get involved at institutions where they might want to make connections. De Sterck added that most people who switch universities after tenure are pursuing specific leadership positions. “If you think that you will be moving, having a really strong research track method is the most important,” he said.

Rouson indicated that the exact opposite is true in industry, as he witnesses frequent movement in Silicon Valley. “If you like moving, then maybe industry is the place to be,” he said. Knepper mentioned that she has been with Intel for 10 years, which is atypically long for an industrial position. She indicated that employees sometimes even return to a previous organization at a later date to advance within the company and obtain new skills.

The speakers agreed that individuals who are looking to diversify their activities should seek out volunteer opportunities. After gaining tenure, De Sterck became involved with SIAM activity groups and conference committees. “If you’re interested in helping to organize things, I think it’s a great way to contribute to the community,” he said. Evans and Tokman echoed this sentiment and urged attendees to volunteer with SIAM, particularly if they are considering future leadership positions in academia or within science in general.

As the session concluded, the panelists reminded everyone that mid-career mathematicians are not done learning and should not be afraid to reach out to their own mentors or establish new connections. It is also important that they continually evaluate their own priorities and change their focus or direction accordingly. “You need to think about which things you’re doing to advance your career, which things you’re doing for your department, and which things you’re doing just to be liked,” Tokman said. “Time is precious, and you can’t spread yourself too thin.” 

Lina Sorg is the managing editor of SIAM News.