How to succeed as an undergraduate researcher

16 minute read

Research is a different experience, with different success criteria, than most undergraduates have experienced before. In contrast to classes and most extracurricular activities, it lacks rigid structure and how to do it well is nebulous at best. The purpose of this post is to flesh out what undergraduate research entails, what succeeding at it looks like, and how to achieve this success.

Typical Undergraduate Research Experience

The exact nature of what you’ll be doing varies by field, but will usually have the same structure. Initially, you’ll work either with a professor or graduate student on one of their existing projects. In the worst case, you’ll be doing something menial for a while that they didn’t want to do. In the best case, you’ll be doing something you find interesting that consistently pushes you just past your current level of knowledge. Either way, you’ll likely meet with this mentor once a week or so, during which you’ll report on what you got done and relevant results, and both of you will figure out what to do for next week. Eventually, if you stick with it and show that you’re effective, you’ll be given more autonomy, ownership of bigger pieces of projects, or even the opportunity to lead a project yourself.

In the first few weeks (or longer) you can expect to be given well-defined tasks. This is partially because you’re not yet familiar with the field and will need specific direction. It’s also because this lets both you and the group figure out how consistently you’ll get work done when classes, etc, get busy. Many undergrads like the idea of doing research, but get little done when there’s competition for their time.

As you get more familiar with the project and research area, your tasks are likely to become less well-specified. This is not because your mentor wants to make things vague, but because fleshing out the details of your work takes time for them. Having you figure out the details yourself is much more efficient. The more you show that you’re capable of accomplishing high-level goals, the easier it is for them.

Once you’ve been with the group for a year or two (if you decide to stay), you’ll likely be the expert regarding some aspect of your project and/or the lab’s operations. For example, you might be the one who wrote a piece of software everyone relies on or the one who collected a certain dataset. At this point, you’re likely to have a decent understanding of the research area the group works in.

Research vs Classes

Because the main responsibilities you’re used to are likely classes, it’s worth pointing out several ways in which research differs.

Need for self-teaching

In classes, it’s generally frowned upon for homework or tests to demand skills or knowledge that weren’t covered in the class materials. In research, there are no class materials. There’s work that needs to get done, and it’s up to you to learn what’s necessary to make it happen. Of course, there’s a limit to what you can teach yourself in a short time frame, so your mentor should keep in mind your current level of knowledge. This is also why it’s important not to try to hide when you don’t know something.

Uncertainty

In classes, success is largely determined by your ability to learn quickly and execute well-defined tasks with high quality. In research, these abilities are important but there’s another quality that matters perhaps even more: Ability to figure out what to work on.

This applies at the level of picking the overall project, figuring what sub-goals will make the project a success, and determining what tasks need to be done on a day-to-day or even hour-to-hour basis to get there. No one hands you a homework-like sheet with numbered deliverables and well-defined specifications. Instead, defining such deliverables and specifications for yourself is much of the challenge.

As an undergrad, your mentor will insulate you from this uncertainty to some extent early on, as described previously. But keep in mind that cultivating the skill of dealing with this uncertainty is much of what it means to grow as a researcher.

Evaluation Criteria: What success looks like (sometimes)

Doing research will get you some combination of:

  • Technical skills associated with the field in which you’re working
  • Understanding of the research process and experience doing research
  • Connections in the form of labmates
  • One or more peer-reviewed papers with your name on them
  • A recommendation from a faculty member

What a “successful” undergraduate research experience looks like depends on your personal goals. If your goal is to pursue a PhD or a research-heavy MS, one of the most important outcomes of undergraduate research is the recommendation you (hopefully) end up with at the end. Recommendations factor highly in university acceptance decisions, and while they may take many forms, an example of a good one might be:

Dear Admissions Committee,

I am pleased to offer Sally Smith my highest recommendation. I have known Sally for 3 years as an undergraduate researcher in my group and have been constantly impressed by her accomplishments and drive.

Initially, Sally worked on X project and did Y. This was challenging because Z, but Sally showed tremendous perseverance and initiative and managed to do Q, which was more than we had even asked for and resulted in her being an author on Paper1.

Sally also worked on project X2, … also an author on Paper2.

Most recently, Sally led project X3, and has operated at a level I would expect from a second or third-year graduate student in terms of her knowledge of the area, ability to independently define a problem, and capacity to invent creative solutions to challenging technical problems, all with only high-level guidance from me. We have submitted a manuscript based on her work to <top venue in the field> and I am optimistic about its chances of acceptance.

In short, Sally is one of the 2-3 best students I have worked with in the past ten years. The others on par with her went to MIT and Stanford. I have offered to have her keep working with me as a PhD student, but I suspect she will both be accepted at and choose to go to one of the premiere institutions in our field. If you have any questions, feel free to contact me at XXXX.

Assuming the papers were published in top venues (and Sally’s first-author paper actually does have a good chance of being accepted), Sally will probably get into almost every school she applies to. Again, this exact structure/set of accomplishments isn’t the only one that can be compelling, but it’s extremely helpful having authorship on papers (especially the first-author one), explicit statement that she’s one of the best in a long time / as good as other students who are now in top schools, and concrete evidence of her positive qualities.

Some non-obvious aspects worth highlighting:

  • Sally worked in the group a long time. She may not have known anything at the start, and nothing indicates she did. What mattered is that she worked hard and showed a lot of initiative early on. Only later did she build up the expertise to be on par with the grad students.
  • There is little chance the professor remembers all of this detail about what Sally did over the past three years. Sally probably kept a meticulous list and gave it to the professor when she asked for a recommendation.
  • The number of papers Sally has is largely out of her control, since it depends on the publishing productivity of her mentor(s). Admissions committees know this, so don’t worry too much about it. What’s most impressive is the last part where she’s writing a first-author paper as an undergrad. As a side note, she almost certainly had to stay and do research over at least one summer to do this.

Dos and Don’ts

Individual work

Don’t: When encountering an unforeseen obstacle, do nothing. When unsure what to do, do nothing.
Do: When encountering an unforeseen obstacle, figure out a way around it or make as much progress as possible with it there. When unsure what to do, ask for clarification or define and execute tasks consistent with your overall goal.

Don’t: Give up when you don’t know how to do something.
Do: Learn what you need to accomplish your tasks. If this isn’t doable, communicate this to your mentor soon enough that you’ll have time to work on whatever revised goals you and they come up with.

Don’t: Require your mentor to figure out the details of what you’re doing.
Do: Learn enough about the project and tools you’re working with that you can figure out the details of what to accomplish and how on your own. This won’t be possible early on, but will become increasingly doable as you gain experience.

Don’t: Blindly complete only the tasks you’re assigned.
Do: Learn more about the research area and project than what’s immediately required to get your work done. This will help you mature as a researcher, decide how much you like the field, and gain valuable context that can help you navigate uncertainty. Note that this learning will take time and just getting comfortable with your work for a while can be a good strategy. Also note that going beyond the call of duty and showing initiative is one of the more effective ways to impress your mentor and/or professor.

Don’t: Treat research like a personal side project that can always be neglected.
Do: Treat research like a part time job demanding serious commitment. You might not want to do this when you first join a group and aren’t sure if you’ll like it, but it will become necessary once you really want to excel within the group. There will be a small number of weeks each semester that are especially busy and most mentor’s won’t expect you to get much done, but this should only be a few weeks—if you get nothing done in any week with, e.g., a test or project deadline, you won’t get anything done after about week four of the semester.

Don’t: Try to keep track of everything in your head.
Do: Take notes on what you’re supposed to do, what results you’ve gotten, interesting ideas or concepts you hear in presentations, or anything else that’s relevant. Forgetting you were supposed to do something is not a valid excuse, and having a record of everything you’ve done will help your professor a lot when they need to write a recommendation for you.

Interactions with others

Don’t: Suffer in silence when things aren’t going well.
Do: Actively communicate with your mentor about how things are going, how long your work is taking, and anything else relevant. Any mentor but a seasoned professor will probably not have supervised many people, and so may not have a sense of the difficulty of what they’re asking of you or how to guide you effectively. Closing the feedback loop helps them mentor you well.

Don’t: Ask your mentor for help whenever you encounter a problem.
Do: First try to solve problems on your own. It saves your mentor both time and mental energy, shows that you have the persistence to deal with challenges, and often results in you learning more.

Don’t: Be afraid to reach out when a quick clarification email could save you hours of frustration.
Do: Identify situations when a little bit of input could save you a lot of time, and then get that input. Combined with the previous point, what this means is that you should evaluate whether it’s worth others’ time to help you when encountering a roadblock. This is situational, but roughly amounts to considering whether it will save time overall when you weight your mentor’s time more heavily than your own.

Don’t: Always work in isolation.
Do: Get to know the research group. This will provide you with both accelerated learning about the research area and get you a network of others in the lab who can help you (especially with using the group’s tools, such as GPUs). Labmates can also help with grad school or job applications, explaining class material, giving advice about classes, etc. One way to get to know the group is to come to group meetings. Another is to work in the same lab space as others instead of at home or in the library (which is your only option in many fields). A related corollary here is that you should take opportunities to present your work to others (including outside the group) at poster sessions or other events; this gets you valuable presentation and communication practice that will help in industry or academia.

Don’t: Be late or skip meetings, especially without warning.
Do: Be on time. If you’ll be late or can’t make a meeting, let your mentor know with as much advance warning as you can. Remember, your mentor is also busy with their own research, and they are taking time out of their schedule to help you with yours.

Don’t: Write ineffective emails.
Do: Write concise, clear emails that are easy to read and provide necessary context. Unless you’re in the middle of a rapid, casual email exchange, it’s usually worth going back through your prose to shorten and tighten it. In some situations, bolding important points (esp. key dates and times) can be helpful.

Don’t: Write casual emails when professional emails are required. “Hi John” is not how you address professors unless you already have a relationship with them and they’re clearly okay with you using their first names. Same goes for writing ungrammatical or overly casual prose.
Do: Address professors as, e.g., “Professor Smith”, write concise, grammatical emails, and otherwise convey that you know how to act professional.

Don’t: Hold back questions because you’re afraid of looking dumb.
Do: Ask questions whenever the answers would help you and you aren’t derailing something (e.g., a practice presentation). No one expects you to know much about the field early on. As mentioned above, you’re evaluated much more on what you get done and the initiative, creativity, and work ethic you demonstrate as you do it.

Don’t: Be unpleasant, disrespectful or arrogant.
Do: Be kind, respectful, and humble. This goes for life in general, but is especially advisable in situations wherein someone’s opinion of you and whether they would recommend working with you matters a great deal.

Summary

A great undergrad researcher is tenacious. They work with enthusiasm, take ownership of their piece of the project, and get stuff done. They get up to speed on the project quickly, figure out the details themselves, and find a way through obstacles and uncertainty. They’re responsible, easy to work with, and constantly improving. They eventually make creative contributions to their projects and become like another grad student.

A not-so-great undergraduate researcher doesn’t care about what they’re working on, assigns low priority to research, and does the bare minimum that’s asked of them (or less). They’re flaky, not around the lab often, and improve slowly or not at all.

Of course, no one is perfect, but the more you resemble the first description and the less you resemble the second, the better a job you’re probably doing.

Further Reading

Updated: