Becoming a PI

Transitioning from someone who does research to someone who enables it.


Eleanor C Sayre

Case studies

As a grad student or postdoc, you were primarily responsible for doing research: knowing the literature, collecting data, analyzing data, and writing early drafts of papers. As a principal investigator (PI), your responsibilities shift. Now, you are primarily responsible for creating an environment in which other people can do research. You’re responsible for all the “big picture” stuff: how do these projects fit together? what’s the future direction of the lab? do we have enough data? what kinds of analyses are likely to be fruitful? does this paper work?

Transitioning from doing research to enabling research

The transition from doing research to enabling research is hard. It’s difficult to “let go” as someone else does things more slowly than you can do them yourself, difficult to watch someone else make learning mistakes, difficult to be excited for someone else’s analysis instead of joining the analysis with them. Framing my role as “research enabler” and focusing on research process has been very helpful in making this transition.

In general, new researchers are not very good at designing what data to take or how to analyze it. You should plan to decide on what instruments/protocols to use, and what the analysis methods are to be. Most undergrads are only available for one year or one summer, so while it’s possible that they could read some papers and do some research, they probably won’t be available to do as much writing as you’d like. Similarly, they won’t have strong enough expertise in the field to make logistical choices about maintaining a lab, or strategic ones about what research has already been done, or what research supports an ongoing program. At the faculty member in charge, you are almost certainly (and exclusively) responsible for maintaining appropriate ethics approvals and ensuring that your students have the appropriate certifications in place.

You’ll spend a lot of time massaging disparate projects together to form a narrative, and most of your personal research time will focus on supporting them rather than engaging directly with data. Undergrads are usually not prepared for theory work unless you get a really strong student and keep them for two years. Don’t plan on that, but enjoy them if you can get them.

Morgan wants to study how first-year undergraduates, particularly neurodiverse students, navigate STEM majors. Their department requires a senior thesis for all majors, and two rising seniors have asked if they can work on this project with Morgan. Morgan wants to mentor them to do this kind of research, but is worried about how to get started.

Morgan decides that their students – who need to write separate senior theses – will work on different aspects of this project, but they should share data collection responsibilities so that they can work together.


Publishing papers with student co-authors is fun, especially if you practice generative writing.

Details vary as to appropriate quantity and venue for tenure, but probably you’re looking at about 1 paper per year on average for a SLAC or regional comprehensive. Realistically, it takes me about 2 years for a paper to move from idea to submission, plus one more year to get published. It takes somewhat more time if there are exclusively undergraduate coauthors and somewhat less time if it’s mostly grad students. Since I aim to publish about 10 papers per year, that means I need to have about 20 papers in progress and 10 under review or in press at any given time (on average). If you’re aiming for one per year, that means 2 in progress and 1 in review at any given time. The best time to get started on your tenure file is when you sign your contract; the second best time is now.

Writing papers for publication is a different genre than writing papers for classes. You should expect that your students do not have any experience writing in this genre. They should, of course, write up their analysis and results! They may even be first authors on papers. However, you should plan that you will need to ghost write a lot of the structure of the paper, and that you will need to spend substantial time editing their prose for genre and audience. Helping them through a couple of drafts is great for their professional development, especially if you work with your university writing center to make it better. However, you may not be able to push their professional development enough in the time that they stay in your lab to make a paper ready for publication. Furthermore, if you push them too hard on the writing, they will stop having fun and leave your lab. You should plan that you will be responsible for a lot of writing, particularly in the introduction and discussion sections which relate this work to a broader literature.

Morgan really hopes that their students’ theses can be published! To promote this goal, Morgan finds a conference proceedings with a submission deadline about one month after their students theses must turned in. After checking with the department chair, Morgan tells their students that if they submit papers to this conference, they can all go together. Morgan and their students plan to structure the theses so that they are as close as possible to the conference submission guidelines, so that there’s minimal rewriting necessary.

How much gets published

On a practical basis, about half of my students’ projects never quite make it to publication. Of the remainder, sometimes multiple projects are combined into a paper. In the 6 years before the pandemic, I directly mentored 10 undergraduates who earned paper authorship, collaborated with another 10ish undergraduate co-authors, and co-supervised 20 undergraduates (most of whom are authors on posters or talks but not papers).

Some depressing stats:

  • About half my work isn’t publishable. Common causes: student moves on; data insufficient; analysis flawed; over-encapsulation.
  • About half my publishable work isn’t published within 3 years. Common cause: I ran out of time to write it up. Sometimes it makes it within 5 years, but usually if I’m not submitting within 3 I probably won’t get to it.

Conversely, this means that 10 papers per year is about one quarter of my research effort. This framing makes me happier than “three quarters of my research won’t get published”.

Sidebar: yes, my lab publishes a ton of papers. That’s a consequence of our publishing-centric lab culture, my predilection for having lots of projects and collaborators at once, and the publishing norms of 3-5 authors per paper. You don’t have to do this to be successful.

How do I write a research statement for a research program which involves undergraduates?

A successful research statement for a small liberal arts college (SLAC) or other teaching-centric tenure-stream job is going to build on your prior work, yet be commensurate with the resources available to that school. So rather than think about what’s hot right now, you should think about what interests you and how that might play out with mostly undergraduate researchers.

When I was at a SLAC, I also borrowed time with other people’s graduate students for some projects. In a research-centric department, I pushed my graduate students to work with my collaborators at predominately undergraduate institutions (PUIs). It’s good for the graduate students to learn remote collaboration skills and understand the SLAC world, and it’s good for my collaborators to borrow time with graduate students. Research is a human endeavor, and it’s better when we do it together.

As as advisor, I’m also happy to continue to collaborate with my former students. Before you plan a research program which centers working with your former advisor, you should check really carefully to see how that might interact with your eventual tenure case: some institutions only count work that’s unattached to your graduate or postdoc advisors towards tenure.

How do I frame research with undergraduates to obtain external funding?

Do you need to? Most research plans, particularly those at SLACs, aren’t externally funded. There will be some internal funding for travel and for students, but that’s not very much money. It’s possible to seek external funding – and you’ll probably be expected to try a little – but not required to actually acquire it. If you’re seeking internal money to support undergraduates, you probably want to emphasize how the money will help their professional development. Ask around in your department: what matters locally?

Sanjay is the faculty coordinator for Science Club, an after school activity for elementary school children ages 6-10. Children meet in Science Club, where undergraduate students are responsible for different activity stations, such as building with Lego or writing code or checking on plants in the garden. Sanjay wants more undergraduates to participate in Science Club more regularly, but he’s not sure how to make that happen given his limited budget.

Sanjay could apply for external funding for Science Club, but that seems like a lot of work, and maybe the Dean can provide all he needs.

The most common US national funding program for discipline-based education researchers to apply to is the NSF IUSE program. There’s often a deadline in mid summer and another one in winter. Many grants are collaborative, so you don’t need to worry about applying alone. Also, the NSF has a low funding rate overall, so it is extremely likely that your grant proposals will not be funded. Don’t take it personally. To get started on what IUSE funds, you should read the solicitation and then search the NSF’s website for awards recently made by the program. There are a number of NSF programs which are amenable to education researchers: IUSE, ECR, BCSER, ITEST, STEM+C, DRK12, S-STEM, Noyce, AISL, REU,… You can read up on all of these. There are other national sources of funding, but the NSF is the most common.

If you’re seeking external funding and your lab is primarily staffed by undergraduates, there’s a careful line to walk: you need to communicate that your science is of high quality (a perceived risk for undergraduate-only labs) yet you are broadening participation in research by employing many undergraduates (a decided affordance). However, this element will be tiny in your grant proposal. Focus on the science.

How do I recruit students to my research lab?

You could just ask. Ask the students in your class if they want to join. Walk into your colleague’s class and announce. Ask your current research students if they have friends who want to do this. Ask the academic advisors in your department if they know someone looking for research. Maybe your institution has an undergraduate research recruitment fair? Make a flyer.

Sanjay’s participants in Science Club come from all over the institution, from science majors to education majors; there are also a few technically-minded theater majors. This diversity is a major strength of the Science Club program, and Sanjay should lean into it when he’s looking for students as research partners.

You’ve got to have a menu of choices for encapsulated projects. Undergrads tend to be more excited about the potential impacts of projects than the theories inherent in them. They respond well to research questions rather than discussions of methods. Many undergrads are motivated by their prior teaching experience (usually tutoring) or perceived deficiencies in their prior learning experience. This group responds well to discussions about measuring or improving student learning, retention, or inclusion. Another group is very interested in doing research (not necessarily in education). This group responds well to generalizable research methods – usually computation – and promises of presentations at national meetings or authorship on papers.

Physics graduates need computational skills, but Maria’s departmental curriculum doesn’t cover computation. She would like to add training in computational skills, but her department is unwilling to require a new course for everyone.

Because Maria’s project is about computation in physics, she could recruit students from the physics department or the computer science department, or perhaps from a new data science minor. Don’t be afraid to look broadly for interested students!

Many departments have an expectation that faculty will engage undergraduates in research, but may be skeptical about how research in education is actually disciplinary for students. To help convince your colleagues, you may want to frame your research as more obviously disciplinary: emphasize content understanding over inclusive practices, computational models over qualitative observations, etc.

Back to top