Tips for Undergraduate Researchers

By Mark Chew
(undergrad researcher in the Pister group 2000--)

 These tips are compiled, based on my personal experience as an undergraduate researcher; I'm sure others could add a lot of meaningful insights beyond what I'm recording here. I'm assuming that the reader is already motivated and just needs some direction in becoming a successful, productive undergraduate researcher.

Finding a Group
Starting Out
Wrapping Up

Finding a group

First, figure out the area(s) within EECS in which you would like to do undergraduate research work. If you're not sure, that's OK too. Just go by whatever you're good at or what interests you. Once you've determined one or two areas, start looking for professors whose research interests match yours. Don't worry too much about choosing your area, since a big part of what you'll gain from your undergraduate research work is experience and insight about how research is actually done.

If possible, try to avoid the approach where you walk into a professor's office and say "can I do some research?" Even if the professor takes you in, you're likely to get a pretty extreme form of grunt work to start out with. (going to radio shack to look for parts, for example) If you were a professor, you probably wouldn't assign important tasks to undergraduates you don't know either. Instead, try to do a little reading on what that professor is working on before you talk. This way, you'll have more interesting things to say, and the professor will get the sense that you actually care enough about his work to have read about it beforehand. Back to top

Starting Out

Once you've found a group, the next thing you need to do is figure out what you'll be working on. This is a choice you need to make carefully becuase it will affect the quality of your experience in the year or so to come. You want to be selective, but not overly picky either. Some questions to ask yourself when deciding on a project are: In addition, it's important to have realistic expectations. I'll be honest with you here. Undergraduates can't do a whole lot, especially in EE, because EE is the type of knowledge that builds upon itself. It takes a long time to acquire enough knowledge to be able to do really groundbreaking research, so what you'll probably be doing as an undergraduate is smaller tasks that fit into a graduate student's overall scheme. You can definitely contribute something significant to your professor's group, but you should expect to work hard before this happens.

It's likely that the first project you work on will be a "grunt work" project. Maybe some graduate student wants you to test his chip or buy parts for him. If this happens to you, don't react violently or anything; this is just part of the normal process of assimilating someone new into a group. While you're in the grunt-work stage, make sure to get your stuff done, but also try to get a feel for the entire group. Who's good at what? What is each person working on? If you don't know, ask them! Most graduate students are happy to talk about their work, as long as you don't pester them about it. Also, this will give you a chance to hear the more difficult and interesting projects that need to be done. The more interesting projects you know of, the better your chance of eventually finding one that suits you. Back to top


Rarely does engineering work go according to plan. You're likely to run into more strange obstacles doing a research project than you would doing a class project because there's a lot less structure. No one's tested the project out beforehand to iron out the technical problems. Expect things to go wrong when you're doing research work; if it wasn't difficult, then professors would just pull any random undergraduate walking the halls of Cory to work on their smaller projects.

Deadlines: Since you're probably going to be working on your project during the middle of a semester, you'll be bombarded with homework and project deadlines, not to mention time-demands from your personal life, as well. Because of this, it's going to be easy for your research project to drop down on your priority list to the point where you're in danger of not getting anything meaningful done, the entire semester. Trust me, this can happen to anyone.

My advice in this area is to set a schedule, especially if you know you can become the procrastinator type. Talk with your graduate student about this. Tell them that you work much better when you're under a deadline, and have them help you come up with a reasonable schedule of deadlines that make you work at a comfortable, but not a burn-out-in-3-weeks pace.

Determination: For me, whether I finished my project or not on time came down to a matter of determination. Since you'll be working in a fairly unsupervised atmosphere, it's likely that this will be true for you too. Part of doing research is just saying to yourself, "When there's a will, there's a way - I don't care what it takes to get this is done." And then you pull an all-nighter and do it.

Reference Letters: You're probably going to ask the professor you're working for to write you a reference letter at some point. All I'm going to say is this: if at all possible, ask them after you've accomplished something. If you're a junior right now, plan it out so you'll have enough time to finish something at least moderately significant before you ask for your reference.

Remember that most professors aren't as interested in how hard you work as they are in what you can produce for them. It's just like in the business world. When you feel helplessly stuck on something, remember that it's almost always better to be ugly and produce something than to have a great approach and not come up with anything. Back to top

Wrapping Up

If you're fortunate enough to have successfully completed the design phase of a project (like taping out a chip, for example), you may have the opportunity to publish. Whether this is appropriate will depend on the project; it's best to talk to a graduate student about this because they will have a better feel than you whether your work is even worth publishing. Even if you don't publish, it's important to document your work for the sake of the next person who tries to use what you've done. Back to top

Other Miscellaneous Tips

Learning how to do Research: It's also important to make the most of your experience as an undergraduate reseracher, apart from the project you're working on. For example, try to observe the graduate students. If you're a Berkeley student, you're working with some of the best students around. What do they do when they get stuck? How does one do reserach in a systematic way? If you can, even try to observe the way they think. When they have an idea, how do they talk about it with other people? When and why do they decide to give up and try something else? Thinking about these kinds of things will greatly add to your experience.

Communication: Talk to people! Don't just sit there and work on your project, without interacting with both your professor and the graduate students in the group. I could elaborate on this, but I'll just let you fill in the details yourself, with your own style.

Evaluation: You never want to get into a situation where you're working on a project that you don't believe in anymore. What I mean by that is, if you stop believing that what you're doing is useful or even that you're able to complete it, you're already dead. You'll lose motivation, and you won't have the determination to get past the next difficult stage of the project. My advice would be to constantly evaluate what you're doing to see if the project is both finish-able, and useful. If you're not sure, ask your professor. (Communitate with him!)

Undergraduate research can become a dead end, where you talk a lot and accomplish little. Or it can be an experience that prepares you to be a graduate student (or whatever next step you take in EECS) and helps you understand the field in new ways. Remember, the quality of your undergraduate research experience is up to YOU!

Back to top
If you want further advice, feel free to e-mail me.