Skeptical about undergraduate research

[Miviato]

I am a sophomore at a highly respected university and, as I decide whether to concentrate my remaining academic years with undergraduate research, I must say that I am very skeptical about it. I mean, I see undergrads involved in advanced physics research without having completed basic courses say, in quantum mechanics. Groups whose main research could only be mildly grasped with a graduate degree, and undergraduates doing their honors thesis with the group. I see these undergrads publishing papers and stuff and I can only question myself: how much merit do they actually have? It took an Einstein to develop general relativity and I see undergrads with papers with general relativity listed as 'basic background for the research"...

 

[genericusrnme]

I've wondered this exact same thing myself, I'd also be interested to hear what someone who actually knows about this has to say about it.

 

[mal4mac]

It took a Newton to develop Newton's three laws, but a bright sixteen year old can use them. There are UG textbooks on GR, an expert in GR can guide a UG through a research problem.

 

[homeomorphic]

I don't know about producing anything of value research-wise, but it is good practice for more serious research, at least. I wish I had done undergraduate research. Things would be a little easier now if I had experience with it.

 

[TMFKAN64]

I once recall reading a comment from one chess grandmaster about another. He said "Given the positions, I can see the brilliancy just as easily as he can. My problem is that I can't reach the positions."

Understanding the brilliancy is much *much* easier than having the stroke of genius in the first place.

 

[101physics]

I did undergrad research for two summers, both in the same field, after completing a quantum course and an EM course (EM probably more important for what I worked on) and I think it helped my physics education tremendously and did not need a graduate level degree to pick up the theory. True, I didn't fully grasp some of the background work at the time, but I was able to understand a lot of what was going on, and you have to remember that your mentor guides you through the research in order to make you learn research skills (practical, such as computational techniques, and theoretical, like deciding which data to trust and why, or which avenues to follow and why) as an introduction. I learned data analysis techniques that I still use and, importantly, I realized that I like research. And yes, I got a paper as second author out of it based on my work and input and I both understood the context of the paper and the results it described. I am now about to start grad school in the same field and while I know that I have a *lot* to learn, doing two (three including my senior year thesis) research projects in the same field has given me a headstart. Remember, too, that your supervisor could probably have completed in a week or two what it takes you the summer or semester to work through, so don't kid yourself that you are doing high-level research work. All it is an introduction to physics research and I would encourage everyone to try some undergrad work before you think about applying to grad school, just so you know if you like it (and can cope when things don't go according to plan or when you don't get a nice unambiguous answer at the end of the project) and seem good at it.

Sorry for the ramblings there - not the most comprehensible post ever but i;m the middle of finals study (help! haha)

 

[Nabeshin]

I am a sophomore at a highly respected university and, as I decide whether to concentrate my remaining academic years with undergraduate research, I must say that I am very skeptical about it. I mean, I see undergrads involved in advanced physics research without having completed basic courses say, in quantum mechanics. Groups whose main research could only be mildly grasped with a graduate degree, and undergraduates doing their honors thesis with the group. I see these undergrads publishing papers and stuff and I can only question myself: how much merit do they actually have? It took an Einstein to develop general relativity and I see undergrads with papers with general relativity listed as 'basic background for the research"...

I think you have to approach undergraduate research with the mentality that it's not like you're going to be making groundbreaking discoveries, but rather that you're simply learning the process of research, along with perhaps a smattering of useful things. You have to realize that anything you can do, a graduate student could probably do twice as quickly, and a postdoc twice as quickly still, along with of course the range of what you can actually do being limited by your coursework. So it's not like they're bringing you into solve the universe's deep problems.

Now, it's true that a lot of people get involved with undergraduate research before they've taken enough coursework to do what might traditionally be called research, but this does not make them useless. In the case of my own research group, we do numerical relativity, but people who have not taken GR are still useful as they can code and produce visualizations. This is generally true, that there are always small tasks in research groups which don't necessarily require a full grasp of the 'mathematical picture' of the research to see. In my opinion, it really helps to know the theory behind what you're doing, because when you spend 10 hours coding every day, knowing the big picture serves as good motivation. Also, if you go around comparing yourself to someone like Einstein, obviously you're going to feel like an underachiever, which is why such a comparison really isn't beneficial. As has been said, undergraduate research is actually quite essential to learning about research in general, and whether you want to dedicate yourself to continuing similar work for a PhD program.

(This coming from someone with 3yrs of undergrad research)

 

[Choppy]

Research is 10% inspiration and 90% perspiration.

The vast majority of undergraduate contributions fall into one of these two catergories.

 

[ParticleGrl]

The absolutely overwhelming majority of research falls into the tedious grunt work category.

If you are in a theory group, this can be coding up lots of checks to make sure your simulation is working properly, or just setting up dozens or hundreds of runs with the code you've developed, or making lots of plots,etc

If you are in an experimental group, you might be winding coils, fixing a leak checker, doing sample prep, etc. Also, of course, making lots of plots.

You basically don't need to know anything at all about physics to follow sample prep someone teaches you. You don't need to know GR to modify an input file and set some numerical code running. The day-to-day in many physics groups doesn't require a ton of physics knowldedge. There are exceptions, like pen-and-paper theory groups, but these generally won't take undergraduates.

 

[twofish-quant]

I see these undergrads publishing papers and stuff and I can only question myself: how much merit do they actually have?

Lots. Most research consists of tedious boring things that can be done without too much background knowledge. There's a ton of work involved in reducing data, dealing with computer output, yadda, yadda, yadda, and that type of research is perfect for someone getting into the field.

Learning to do research by doing research is important because it gets rid of a lot of misconceptions, namely that most research consists of "eureka" moments, and it gives people experience on how much grunt work is really involved.

It took an Einstein to develop general relativity and I see undergrads with papers with general relativity listed as 'basic background for the research"...

It took Newton decades to come up with calculus. For that matter the development of algebra took several hundred years.

It's not difficult to get a background knowledge of GR that is good enough to do research. For that matter, it's also possible to use GR as a black box, and having black boxes is something that frequently happens among professors.

 

[motx]

I got to work on stuff that was years beyond what my freshman training had me ready for and now my name is printed somewhere in some little chemistry journal because of it.

I certainly didn't make any brilliant discoveries. I certainly didn't even understand the chemistry as well as a person could. I did, however, make a few useful suggestions on how to run our reactions. I also got to see just how slow and tedious research actually is.

The professor I worked for kept track of all of his students after graduation. None of his undergrad research assistants accomplished anything astounding before grad school, but quite a few did go on to do some really interesting work later on down the road.

I do think there was a lot of merit to their work and to my work. It's just not in the way you seem to be thinking.

 

[twofish-quant]

I think you have to approach undergraduate research with the mentality that it's not like you're going to be making groundbreaking discoveries, but rather that you're simply learning the process of research, along with perhaps a smattering of useful things.

You have to approach *any* research with the mentality that you aren't going to be making groundbreaking discoveries, and are just doing the grunt work that has got to be done.

There is a huge amount of luck in research. Most astronomy research consists of "I look in my telescope, and this is what I saw, and what I saw us more or less what I expected to see." And then there is this lucky team that was planning on doing that and what ended up happening was "I look in my telescope, and holy crap, the universe is accelerating, and I got a free trip to Stockholm."

One other reason for getting involved in undergraduate research is that you find out quickly if you really like research. You may find that you hate research, which means that it's a bad idea to spend your life doing it.

You have to realize that anything you can do, a graduate student could probably do twice as quickly, and a postdoc twice as quickly still, along with of course the range of what you can actually do being limited by your coursework.

Or maybe not. It's often the situation that younger people have skills that older people haven't had time to pick up. In my generation, it was computer programming. For people in college nowadays, I'm sure there is some useful skill that could be used in research that older people don't have (and it probably has something to do with youtube or facebook).

 

[hadsed]

Undergraduate research teaches you a lot of things, and some of the things can be exactly what you are thinking of. Most of them aren't, though. However, you'll be surprised at how much an intelligent, well-guided, and encouraged undergraduate can do.

So I started doing research in the second month of my undergraduate career. I'd been a programmer for a long time before college, so I was able to use that as grounds for my advisor to hire me. As a student, something that gets to be really important when starting to do research is that you understand the situation you will be in. It's going to look hopeless (as a freshman, it was an incredible level of hopelessness), you're going to feel like you couldn't possibly learn, know, and understand everything that's going on in the field, and you won't know if anything will even come out of the work you're doing. It's easier not to worry about this as an undergraduate, and to only focus on how well you can learn to learn.

Something important that I learned when working in numerical astrophysics is how to take a computer, figure out how to tell it things, and make it do math to simulate a physical system. This is something that, if you've had no prior exposure to, is very difficult to imagine. In the beginning months of my research, I spent all of my days thinking about how these problems could be solved, what is lurking in the depths of these massive codes, what higher level mathematics is being used that I have no semblance of right now. It was certainly something that, to a curious, research-oriented mind like my own, required long periods of deep thought as well as hours and hours of sitting on a computer reading the 30 tabs of Wikipedia articles I had opened. I was lucky to have an advisor that was extremely patient with my ignorance of even the basic mathematics needed for research. For example, my first mini-project was to find the gradient of a gravitational potential for a dark matter halo, and then to implement by writing some code to calculate the force on a particle in that field. It was a bit weird to me at first considering I had just enrolled in integral calculus for that semester, but luckily I was able to learn. The math method itself wasn't difficult to learn, but if you're like me you'll be dying to understand how everything works, and you'll despise using things with only a superficial understanding. However, at some point you will realize that you must take things as they are, because you aren't ready to understand it quite yet.

When I moved on to work in a condensed matter group, a year later, I had a much deeper understanding of the physics I was studying. I was (and am) able to try and start working on my own implementations for solving physical problems (and understand and implement numerical algorithms developed by others to my specific problem). After a year of being in the research state-of-mind, I feel that my knowledge and understanding of physics and the related mathematics is much better than I could ever have gotten from sitting at home and doing all my homework properly. There is the added bonus that you become more and more effective at research, as long as you're given some meaningful work to do (of course, I do my fair share of running simulations and writing data analysis scripts, but I'm also heavily involved in developing new approaches to solving the main problems we need to solve). I learn much better when I'm given a problem that I have no idea how to solve, because I'll go find all of the information I need, read and learn it, and then solve the problem. In this way, anything and everything is relevant because it may hold the key to solving my particular problem, so I feel that I learn more material more effectively, with a deeper understanding. There is a slight motivation factor when you consider that your research project could have big outcomes, something that a homework or even a test will never do for you.

I know we are talking about the merits of this research, but I feel compelled to mention that there are other reasons you should do undergraduate research as well. If you are wanting to get a graduate degree, it takes a lot of stress off your first years when you know what to expect. As a graduate student (not from personal experience, only second hand), you won't immediately begin your thesis project. It might be random walk-ish in terms of projects that you work on before you get to a point where you can choose a good thesis topic. Learning to deal with failure in research is a huge part in progressing without going insane, and if you've dealt with this while having some personal guidance from a professor, it will be much easier to be independent in graduate school. It takes a lot to get a Ph.D, and having an understanding of what it will be like before you tackle it may mean the difference between you dropping out after you finish up your master's courses or plowing through to the dissertation defense.

The other thing to mention is that these research problems don't have any time limits; you're not meant to finish it in a two hour class period. Actually, the correct thing to say is that it wasn't designed to be done in two hours (because, of course, it wasn't designed, and that is why it's so exciting!). These problems, as I mentioned before, were at the center of my mind most of the time. It was on my mind while I was walking, eating, driving, sitting in class, showering, etc. This is the typical way you think of research problems I think, and it's good to get used to this. It may sound dreadful to some, but it isn't something that you must force yourself to do; you will feel the compulsion to solve the problem on your own, and your brain will constantly be working towards it.

Some people thrive on this sort of thing, and it'll push them to learn at an incredible pace, one that is unmatched by any classroom learning. This is the largest benefit, to me, of doing research as an undergraduate. Whether you have the mind/personality for research or not will become clear to you as soon as you start doing it. If it's not for you, you're better off knowing now. If you do have it in you, you've built a very good intellectual and psychological foundation for your future career as a researcher.


[Just for transparency/honesty, I'm finishing up my second year as an undergraduate, so don't hold me to my word on graduate life experiences or teaching methodologies, they're only what I've read or heard (or, in the latter case, my opinion).]

 

[hadsed]

Or maybe not. It's often the situation that younger people have skills that older people haven't had time to pick up. In my generation, it was computer programming. For people in college nowadays, I'm sure there is some useful skill that could be used in research that older people don't have (and it probably has something to do with youtube or facebook).

Yeah. Also, I think it's still computer programming and related skills. In the more hardcore numerical fields, it's not usually the case where the senior researcher doesn't know how to write parallel code, use newer optimized libraries, different languages, data analysis packages, etc. but there are a lot of fields where these problems start to come up and where these things might be useful. For instance, one of my research advisors doesn't know C++, but it's not so bad because he knows FORTRAN really well. There's a guy in a group we collaborate, the principle investigator actually, who doesn't know anything numerical and just gets postdocs to do his dirty work.

But yeah, there are a lot of other things besides that too. For example, I can find something on Google 10,000x faster than my advisors can. Or if they need some integral, I've already computed it on Wolfram Alpha by the time they've gotten their table of integrals book out. Or definition of some SL function, and so on. So search seems to be one of the big things that is pretty natural for the youngins to be able to do easily.

 

[jtbell]

The absolutely overwhelming majority of research falls into the tedious grunt work category.

And it will very likely turn out to be insignificant in the grand scheme of science.

See Sturgeon's Law.

Which isn't to say that it shouldn't be done... because it's very difficult, if not impossible, to predict where the next "really significant things" are going to come from.

 

[twofish-quant]

But yeah, there are a lot of other things besides that too. For example, I can find something on Google 10,000x faster than my advisors can. Or if they need some integral, I've already computed it on Wolfram Alpha by the time they've gotten their table of integrals book out. Or definition of some SL function, and so on. So search seems to be one of the big things that is pretty natural for the youngins to be able to do easily.

Yup, and as I get older I'm getting more sympathy for old people. When I was younger I was wondering why the older professors didn't learn all of this new technology, and now I'm getting old, I'm realizing that the reason why is simply lack of spare time. My kids are much more adept at using skype and facebook than I am, and its not so much that I can't use those tools, but rather 1) I don't have the time to learn and 2) all of their friends are online. Mine aren't.

One thing that people in college don't realize is how valuable spare time is.

 

[genericusrnme]

And it will very likely turn out to be insignificant in the grand scheme of science.

See Sturgeon's Law.

Which isn't to say that it shouldn't be done... because it's very difficult, if not impossible, to predict where the next "really significant things" are going to come from.

This is the single greatest law I've ever read about!

 

[Andy Resnick]

I am a sophomore at a highly respected university and, as I decide whether to concentrate my remaining academic years with undergraduate research, I must say that I am very skeptical about it. I mean, I see undergrads involved in advanced physics research without having completed basic courses say, in quantum mechanics. Groups whose main research could only be mildly grasped with a graduate degree, and undergraduates doing their honors thesis with the group. I see these undergrads publishing papers and stuff and I can only question myself: how much merit do they actually have? It took an Einstein to develop general relativity and I see undergrads with papers with general relativity listed as 'basic background for the research"...

I am currently hosting 4 undergrads (in addition to grad students) in my lab, and they range from sophomores to seniors. They have a range of skill levels, understanding about what research *is*, and various levels of understanding about the overall research program.

You say you are 'skeptical about [research]', but you haven't yet actually participated in a research project. I recommend that you don't make judgements about what you think undergraduate research is.