A roadmap for undergrads in science


I’ve found myself recently giving out some advice about the US grad school application process. It’s already October, so the `grad school train’ is already making its way out of the station. (All aboard who’s coming aboard!) Why didn’t I say something earlier? I did.

But that was in July. The path to grad school really begins the moment you set foot in your freshman dorm. Here are my own thoughts to pass on to the next generation of science undergrads.

Fig 1. No matter what your goal is after your Bachelor’s degree, the road ahead probaly isn’t well defined. Adapted from a popular photo on the Internet.


First of all, if you don’t know what you want to do with your life: do what it takes find out now. If you think you might want to be a scientist, then work as a research assistant in a lab. If you think you might want to be an investment banker, then go get an internship. Even if it doesn’t work out, then at least you know that a certain path isn’t right for you and you can continue looking.

If you want to be a scientist and are considering graduate school en route to a research career, then read on. (If you want to be an I-banker, then go do that for a while and come back when you’re sick of it. 🙂 )

By the way, it’s become increasingly popular for private universities to offer fifth year master’s degrees to undergrads. There are a few people for whom this is the right choice, but for the rest it’s really a scam to wrangle another $40k+ from your parents while you try to figure out what you want to do after graduation.

Nobody will tell you what to do…

Aside from acne and senior prom, high school was straightforward. I didn’t say it was easy, but at least the path ahead of you was clear: finish your homework, study for exams, and do whatever your college counselor says. (Extra credit if you are involved in science research in anyway.)

But university isn’t like that. For the most part, it’s up to you which courses you take and what you do with the rest of your time. Further, unless you go to a school whose name ends in “… Institute of Technology,” most of your fellow classmates won’t be going into research careers in science. “Following the crowd” may not be the best thing for your future scientific career.

In many respects you are on your own to make about how you want to be trained as a scientist. It’s up to you choose what field(s) you want to specialize in, which courses to take, whom you do research with, and in general how to take advantage of the resources available to you.

That’s why it’s really important to have a clue about what’s going on. Otherwise, you run the risk of realizing in your senior year that you don’t have the right background to apply to postraduate programs in astro-psycho-physical-bio-chemist, or whatever. Part of the game is figuring out the rules.

… unless you ask.

So if nobody is there lighting up the path between freshman year and picking a grad school, how do you figure it all out? All you have to do is ask.

Take the initiative to talk to faculty and graduate students. They were undergrads once, and they’re usually very happy to give advice to undergrads who want to follow in their footsteps.

Yes, it may be intimidating the first time, but you won’t get anywhere if you’re scared of knocking on a professor’s door.

Also, talk to older undergrads in your field. Check if your university has a Society of Physics Students (or the equivalent for your field) that hosts social events.

The first commandment of being a good undergraduate is to ask for advice from those around you. The more advice the better.

No Peacocks

This is a corrolary of the first commandment: ask questions in class. This was the best single piece of advice I ever received in any discussion section (“example class”) during my undergrad years.

As my freshman E&M TA put it, there is some strange `Peacock Syndrome’ among physics (and other `techie’) students. They seem to be obsessed about the image that they know what’s going on, even if this comes at the expense of actually knowing what’s going on.

To this extent, a physics student won’t ask questions in class. It would be terribly embarassing if one asked a stupid question, because then everyone would know how stupid one is. So instead, physics students sit quietly and preen their feathers so they can cultivate their image.

This is stupid and harmful.

First of all, this is stupid because there’s a professor/TA who’s there to answer questions about physics. If you’re not going to ask questions and engage in discussions, then it’s a waste of time for everyone. You might as well quit university and learn physics by reading textbooks and watching online videos.

Secondly, it’s only going to get worse. This is like meeting someone and forgetting their name. You can ask again right away, or you can wait until you get to know each other more and then realize that now it’s really inappropriate to be asking their name. So suck up your pride and ask.

Thirdly, Critical discussion is a different mode of thinking than passively listening to a lecture. You will learn more when you’re processing information in real time.

Science is all about asking questions. Learning how to ask good questions (and inevitably asking one or two embarassing ones) is important.

Your classmates are not your competitors

Perhaps the whole `peacock syndrome’ is an outgrowth of a sense of machismo associated with physics (or engineering, etc). Some students may feel like they’re out to `do better’ than their classmates, since this strategy may have worked in high school where it’s easy to focus on having the best GPA.

This is also bad. Modern science is based on collaboration. There’s a reason why faculty encourage students to work together on problem sets — this is training to be able to communicate scientific ideas clearly. (Also it means they don’t have to repeat the answer to the same questions over and over during office hours.)

If you want to be really Machiavellian about it (and I don’t think you should), then you need to understand that the people you are `competing’ with for grad school positions aren’t just your classmates, but every other physics student in the world. And many of those students have figured out how to study together so that they all benefit. By not collaborating with your classmates, you’re only making it harder for yourself to be competitive with the rest of the world.

But, like I said, the competitive point of view isn’t a healthy one. Instead, understand that your classmates are resources. They offer different perspectives about how to understand the material.

If you haven’t figured all this out yet, then trust me. Work with people in a study group. Yes, even if you’re skeptical that they can teach you anything. The worst case is that you’ll be the one explaining things to them — and you’ll be surprised about how much you’ll learn just by doing this.

Do your problem sets

Maybe high school math(s) and physics was really easy for you. You just had to read a chapter and the concepts were straightforward. Or maybe you just rederived everything from first principles on the exam. Good for you.

But things are different now. You’ll be taking courses at your level, and they’re meant to be challenging. (If they’re not, then you’re in the wrong class.) If had it easy in high school, it may be difficult to develop good habits at university. People who have struggled at some point, however, know that working through problems solidifies one’s understanding.

Yes, you’re busy. But you’re not really learning anything if you’re not applying it. The skill set you want to develop isn’t internalizing knowledge, it’s doing something with it. If you’re short on time, then only do the problems that you can’t do. That’s not me being facetious;  you’ll learn the most from banging your head on hard problems.

If you want to be a super-good student, then start a notebook of solved problems. Go through interesting textbooks or a book of physics problems and jot down your solutions nicely and legibly. Maintain this habit. When you’re doing more involved research problems down the road, you’ll be happy to have volumes of worked out examples in your own words that you can refer to.

Some suggestions of problem books: 200 Puzzling Physics Problems for freshmen and high school students, “Problems and Solutions on ___” by Lim for upperclassmen. For more advanced students, Cheng and Li’s problems and solutions book (caution: I’ve caught a few unlisted typos).

Do research

If you want to be a scientist (and often even if you don’t), then you really must participate in undergraduate research. The more the better.

Fig 2. Image adapted with permission from “Piled Higher and Deeper” by Jorge Cham. Comic number 508. Don’t worry, undergraduate research isn’t really like this… most of the time.
As a high school senior, I never understood why universities would promote their undergraduate research programs. The point of going to college was to take classes and so that I could do research as a grad student, right? Wrong.

Based on my humble observations of postgrad physics programs in the US, solid undergraduate research experience is the dividing line between students who get into top programs in their field and those that do not. It makes sense. An applicant with lots of research experience understands what’s expected of a graduate student and has a head start in adapting to that environment. This is a stronger indication of grad student potential than grades or any standardized test.

If that doesn’t convince you, then here’s a short list of some of the reasons why undergrad research should be a centerpiece of your university years:

  • Doing problem sets and lab classes is not the same as doing science. UG research introduces you to the actual scientific process.
  • UG research helps you test-drive a particular type of science. How else would you know if you would like to ultimately work on solid state physics, or astro, or (fill in the blank)?
  • Working in a research group naturally puts you under the care of a grad student / postdoc / faculty mentor. Their advice for your academic career is valuable.
  • Research students usually get fantastic grad school letters of recommendations from advisers who can write substatially about a student’s scientific potential.
  • If you’re able to take a project a bit further, the experience of publishing a paper or giving a conference talk is priceless

By the way, your first UG research experience probably won’t be perfect. In fact, during your undergrad career it’s more likely that you’ll be a lab jockey rather than a wunderkind. That’s okay. The most important thing is to understand how research works: the creative process, troubleshooting, the way questions are posed and the different approaches to a solution.

This means that it’s okay to do experimental research in condensed matter physics even if you ultimately decide you want to do particle theory.

For those at liberal arts colleges without much research support, always keep an eye out for external research opportunities at national labs and nearby research universities. There are plenty of options out there, you just have to look for them.

Have mentors

I’ve mentioned seeking advice already. But go beyond that and seek mentors. Academia is still modeled on this Middle Age system of apprenticeship where faculty pass on skills and a way of thinking to their students. This is most explicit in the relationship between a professor and graduate student, but this doesn’t mean undergrads can’t benefit from this as well — especially those who are engaged in research with a faculty member.

Mentors provide more than advice about what the next step in your career is. They are role models to try to emulate, provide a structure of support, and pass on `good physics habits’ along the way. Mentors needn’t be faculty members — department staff, postgraduate students, and advanced undergrads all have useful things to share with students who are just embarking on their scientific careers.

I almost certainly wouldn’t have made it to postgrad if not for the support of my mentors. It’s not so much that they directly taught me a set of skills that solved all of my problems (though they did teach me lots of great physics); but rather they helped me though the maze of undergrad responsibilities.

It’s commonly accepted that college friends will really shape your social personality. What’s less known is that your university mentors are the people with whom you develop your scientific personality.

Stay on top of things

Like I said: nobody will tell you what to do [without you taking the initiative to find out]. Recently the ETS cancelled their December Physics GRE test, causing some significant consternation among at least a few of my friends. I don’t want to point fingers, but grad school applications are important—you have to stay on top of things. “But I didn’t know” isn’t an excuse. There are no do-overs.

Academia is a fantastic field to work in — but don’t underestimate its potential to be cut-throat and competitive at times. If you’re lazy doing a project, then you’ll get scooped. No consolation prize.

Be responsible for yourself. (It’s not always cut-throat, but you can’t go about asking for sympathy.)

Fake it ’till you make it

This is one of my favourite Jazz quotes. (I think it’s an old Jazz quote.) In many ways figuring out the odds and ends of your scientific career is a lot like improv. In many ways it helps to apply this philosophy to science.

I do not mean fabricating data or cheating on coursework! What I mean is developing healthy scientific habits, even though they might not be clicking for you as an undergraduate. One manifestation of this is undergraduate research, which I’ve discussed previously.

Another important example is attending colloquia. For those not familiar, departmental colloqiua are seminars given to a broad audience of physicists. Thus a talk on high temperature superconductors needs to be sufficiently accessible for a physicist in a different discipline, say astrophysics, to follow along. These are a great way to get exposed to current research topics and see what’s going on at the forefront of physics. You might not really need to know about these things, and you probably won’t understand every detail — but it does introduce you to the scientific community at large. (Also, I always thought it was funny to watch how my professors interacted with one another.)

An extension of this is to attend seminars in your field. These are more specialized talks on current research, so they may be even more difficult to follow along. With a bit of preparation (by talking to a graduate student, for example), you can at least pick up tiny bits and pieces of whats going on. You start out learning nothing. Then you learn just a little bit. Then a little bit more. And so forth until your education catches up and you begin to understand more at an exponential rate. The purpose of doing this is to familiarize yourself first with jargon, then with broad ideas and approaches, and then finally with the current status of your field. If you want to be someone who “knows what’s going on” as a first year grad student, this is the way to do it.

Finally, one of the habits that I think took me very far was talking about particle physics with a group of friends. A group of five of us were very ambitious and took courses that were probably slightly beyond our comprehension; but we would eagerly chat about the latest seminar or controversial paper. We started out by just spouting terribly stupid nonsense and making silly jokes about academic jargon. But as we started learning more and more, we would bounce ideas off one another to check our understanding of material. As we branched off into our own research directions, we would talk about our understanding of our fields and explain ideas to one another. I think we learned a lot from one another in our last year — but the most valuable thing we each took away was a shared self-perpetuating enthusiasm for physics. (A shout out to Chris, the Dans, Josh, and Naresh!)

Develop skills

Use your undergraduate years to develop a set of skills that may either be essential in grad school, or that might otherwise make you unique. These skills may include a strong background in certain areas of mathematics, a knack for electronics, or even knowing how to make a nice figure for a paper. It’s a bit hard to predict what will be important down the road, so don’t be afraid to branch off and spend a bit of time doing things outside the mainstream. For example, proficiency in a foreign language might turn out to be really helpful at some point in your career.

There are a few things, though, that I think everyonoe should know by the time they graduate:

  • Know how to program. The exact language doesn’t really matter, but understanding how to properly write a program is a prerequisite in many fields (and is very helpful in the rest). It’s not hard to pick up another programming language after you properly understand one. Popular langauges in science include C/C++ and FORTRAN.
  • Know how to prepare a scientific document. Mainly this means understanding how to use LaTeX and GNUplot, since you’ll learn a lot about the scientific writing process when you write your first paper.
  • Know how to give a good presentation. If you have stage fright, get over it before you’re a grad student. And try to get as much experience as you can giving talks about your research; both to a scientific and a lay audience.
  • Learn how to cook to survive. If you’re living on a posh dining hall meal plan, you might want to be sure to learn some home cooking recipes froom home before you go to grad school. Those PhD stipends won’t last long if you’re eating out every day.

Stick out, take calculated risks

Depending on how you crunch the numbers, the chances of a high school physics student becoming a theoretical physicist is about the same as the chances of a high school basketball player becoming an NBA player. It helps, sometimes, to stick out a bit.

For those who stick out because of their pure genius, good for you. For the rest of us, this part usually requires a lot of hard work and a lot of commitment. Sometimes it also means taking calculated risks.

In my case, my big risk was taking two years out of the `usual’ career path to do masters degrees in the United Kingdom. I think it’s worked out very well, but I’m also aware that I’ll be two years older when I start my PhD (which can be an issue for those planning families and such). Other risks include deciding where (and with whom) you do your PhD, or what kinds of research projects you decide to become involved in.

Not all risks are worth taking, but don’t be afraid to think different(ly). Those are the people who end up getting noticed.

Want it

As a final word, you can lead a horse to water, but you can’t make it drink. The most important piece of advice I can give is that you really, really have to want to do what you’re doing. It takes a while to figure out what it is, exactly, that you’re doing — but once you figure that out, follow it to what you enjoy.


15 Responses to “A roadmap for undergrads in science”

  1. Wow, what an amazing post. I can guarentee that every undergrad who reads it will go away with something valuable.

    Thanks for such an excellent blog.

  2. 2 alphabet

    What happened to your friends when the December GRE was cancelled? Were they allowed to submit their app witout a score?

  3. They rushed to take the November exam. I’ve heard urban legends of students who have applied without GRE scores and who were `so good’ that the departments accepted them anyway, but these are all “friend of a friend” stories.

  4. 4 Gribin

    If the chances of a high school physics student becoming a theoretical physicist are comparable to the chances of a high school basketball player joining the NBA, then how much have your chances advanced now that you succeeded as an undergrad and have been accepted into a good graduate program? What percentage of Berkeley physics grad students go on to academic careers? If that doesn’t work out after years of intense study, what’s next?

  5. Allow me to offer one of my favourite quotes:

    `If the worst thing that can happen [studying theoretical physics] is that I drop out and become a millionaire on Wall Street, then things aren’t so bad.’ -Nav S. (who is otherwise a notorious pessimist; c. 2005).

    It’s a bit saccharin, but at the end of the day we (scientists, grad students, whatever) are doing something that we really love, and that’s the main motivation. I suspect that the people who are the most passionate about their field are the same people who can persevere through the setbacks (and there are always setbacks) and finally get to that faculty position.

    That’s the way the system is set up at least. It’s not perfect. There are good scientists who get lured away to industry. But, as N.S. said, the PhD is a stamp that says you are highly skilled (in a general way: problem solving, perseverance, etc). Unlike high school NBA prospects, a PhD in physics will go a long way no matter what your ‘plan b’ is.

    I don’t really know what percentage of ____ students go on to academic careers (for whatever ___), and I don’t think this is necessarily a meaningful statistic. Whether or not a PhD student decides to continue in academia doesn’t necessarily correlate with whether they’re good enough to do so. Try asking faculty in your department if they know of recent grads who could have gone on to very good academic careers but instead opted for another path. It might be more common than you expected.

    For me personally, I’ve always been interested in teaching or writing. When i was a child I dreamt of being an author. As an undergrad I enjoyed doing high school outreach programs (I’m working on a template for a particle physics master class) and was tempted to do Teach for America. Those options are still on the table if physics doesn’t work out. But I don’t like to think along these lines. One of my advisers told me that once you’re sure that physics is right for you, then the best course of action is to go for broke and apply yourself 110%.

    To quote Andre Linde (paraphrasing):
    “A few years ago the department asked me to talk to prospective physics majors. So I went and I told them: `If there’s anything, *anything* at all that you can imagine yourself doing other than physics — then you should do that instead. Otherwise, you should be a physicist.’ They never asked me to talk to undergraduates again.”

  6. 6 kni

    Hi.. thanks for this great post.. regarding the collaboration bit.. in high school.. always believed that not studying in groups would be better for me so that I can learn to think on my own, rather than relying on others.. which will serve me well during the exams.. is this still true in college? How can I benefit from study groups yet avoid the over reliance on others?

  7. Hi kni!

    My advice would be to approach the collaboration as a chance to fortify your own understanding in addition to learning how your classmates understand things. Note that it’s important to be on both sides — you want to take the opportunity to explain your point of view as well as absorbing others’ points of view. This almost always leads to a better understanding for everyone involved.

    Don’t worry about over-reliance on others. A good litmus test is to see if you can reproduce the ideas (e.g. deriving key equations in a problem set) on your own afterward. If you can’t, then you weren’t interacting enough and were too passive.

    Hope that helps,

  8. 8 kni

    Thanks for your reply!

    To be honest, I’ve never been on a study group before.. Have been kinda a lone wolf in high school.. So how does it actually work (it’s actually kinda embarrassing to ask about this as someone who’s going to attend college)?.. Do I finish all my reading, problem sets and ‘thinking’ on my own before joining up with a study group to discuss what I understand (my ‘own’ picture) and ask about what I don’t get? Does it work like this?

  9. 🙂

    Comment threads like this make me want to TA first year students this fall. I had a really influential TA in my second term that really emphasized approaching undergrad as a preparation for becoming physicists.

    To answer your question, it’s really up to you. I’ve been in study groups where we all sat down to attempt the problems on our own and only started to discuss when someone got confused. Perhaps a slightly better model is for everyone to have looked and thought broadly about the problem set ahead of time, then working out the details together. Sometimes you can just do as much of the problem set as you can by yourself and then ask a friend for some help for the remainder (this usually involves offering to buy a coffee).

    One should also take advantage of faculty and TA office hours to ask questions when everyone is stumped.

    The important thing is to be proactive. Start problem sets early, even if you’re just taking a cursory look at them. Treat them like research problems where you can’t rely on begging a TA for the solution the night before the homework is due: use the entire week to talk through bits and pieces with colleagues. You don’t even need to have a formal study session, just asking a few questions after class can be a big help. You *really* just have to make sure to leave time for this process to occur. If you waste even half of the week, you’ll be under time pressure to write down your solutions, which prevents you from really thinking through the problem effectively.

    It is prudent to keep up with the course in terms of textbook reading. Depending on the style of the lecturer this may or may not be a requirement, but it’s always prudent. If you wanted to be a superstar, you’d try to read a bit ahead of class so you could ask intelligent questions at points where you were confused in the book. … but this may also mean you have too much time on your hands and maybe should take a more challenging course load. 🙂

  10. 10 kni

    Thanks for your advice..

    Also, I’m heading to a liberal arts college.. what do you think are the advantages/disadvantages in terms of trying to get into top graduate schools? We don’t have nobel-laureate-superstar professors here to get recommendations but the professors are here for us..

  11. It’s probably generally true that faculty at liberal arts colleges focus more on students’ education, but as you note this comes at a cost of some of the perks of research institutions.

    At the end of the day, students from liberal arts colleges get into many top-tier grad programs. (Students from places like Reed seem to do exceptionally well.) But that being said, you need to always keep in mind that you’ll be competing for positions with students who do have the benefits of going to research universities.

    As such, you should really make an effort to ride the advantages and make up for the disadvantages of where you are (whether it’s a liberal arts school, a research university, etc).

    Grad schools are looking for students who have shown that they’re able and ready to do research. Some quick thoughts on this:

    1. Get to know faculty in your department. Pique their brains for advice about your undergrad career and your plans for grad school. Do well in their courses and make sure they remember when it comes time to write recommendations.

    2. DO RESEARCH. The biggest advantage that students at research schools have over liberal arts schools is that they have ample opportunity to do undergraduate research during the year and over the summer at their own institution. On the other hand, there are plenty of summer undergraduate positions at national labs for students who don’t have these opportunities — but it’s up to you to keep an eye out and apply for these things. (For particle physicists, there are good opportunities through the DoE SULI program, the CERN summer student program, and even the Perimeter Institute has an undergrad summer program. Additionally you can e-mail schools like Caltech to see if they have positions for external undergrads.)

    3. Have a stellar academic record. There are two things that can work against you at a liberal arts school: (a) you might not have as much access to graduate-level courses, (b) fewer of your colleagues may be interested in an academic career. You’ll need to always remind yourself that ‘the bar’ isn’t necessarily set by your immediate peers, but by all of the students across the country who are hoping to do a PhD. If you don’t have access to a graduate level QFT course, then at least make sure you’ve taken all the relevant courses in the physics department and as many relevant courses from the math department. See if faculty will let you do independent study reading courses in advanced topics.

    As a final note, don’t get hung up about the differences between other universities and your own. The quality of your education is primarily under your control, not the university that you go to.


  12. 12 kni


    Thank you for your reply. I’ll watch out for those things when I get to college.. Also, I am a non US citizen.. so a lot of the summer research programs are shut out for me.. There are still summer programs available to me.. but few and far between.. ultra competitive to get in..

  13. 13 kni

    Also, grad school admissions are all about GPA, GRE, Recommendation and research right? ECs do not count at all? How about ECs which are related to science? Like a science club? How about being a teaching assistant/peer helper?

  14. Hi Kni — research programs become a bit more difficult if you are not a citizen, but they are not impossible. You’ll need to take more initiative to contact departments to see what opportunities are available. For example, my undergrad university had a little known scholarship for minorities which was used to bring international students on campus to do summer research. Also consider programs at non-US labs, especially CERN (where it appears to actually be easier to get in as a non-US citizen).

    For grad school admissions, EC’s generally don’t count for anything. It’s nice to show some teaching or science outreach interests, but that’s really not why they’re going to accept you. (Though this does help for some US fellowships, namely the NSF.) Grad schools care about your potential as a research student. Most research universities will *always* take a better research student, even if that means that that student isn’t a very good teacher.


  15. 15 kni


    Thanks for your comments. I’ve looked through the Stanford website. But I don’t see any research programs for undergrads.. I only see those for Cardinals.. Maybe I didn’t look hard enough..

    Were you that sure that you’re headed to graduate school when you started college?

%d bloggers like this: