Let us eat cake!

Wednesday, May 5, 2010

Wrong. Just wrong.

Sorry I haven't blogged in a while. I just have to say that I vehemently disagree with this call for less emphasis on research in undergraduate institutions. Now, certainly the writer is correct in that there are certain obstacles to research in undergraduate institutions. Believe it or not, PUI faculty are well aware of this. No, really! And certainly it will be true that PUI faculty will not bring in the same funds as their colleagues at research universities. However, this does not mean that PUIs should be abandoning the best means of teaching science. We'll have to be creative, we'll have to look for approaches that work in our situations, but I cannot think of a better way to teach science than to do science with a student.

Also, I would note this: When I came to my job at Cal Poly, I new that I would have to recalibrate my expectations as time went on. To my delight, the science that I am doing with my students is far better than anything I was thinking of doing when I started. And we continue to come up with even cooler things as time goes on.

Tuesday, March 2, 2010

Do we send too many students to grad school?

A provocative article in Scientific American argues that we aren't under-producing Ph.D.'s, but we may be producing them in a manner that is poorly aligned with the job market. This seems relevant to the question of what constitutes success for an undergraduate research program: Is an undergraduate research program simply the beginning of a "pipeline" (I truly hate that metaphor) that ultimately feeds into the higher tiers of academia, or is it simply a way to teach a combination of technical skills and problem-solving skills in an environment where the answers aren't known? I lean quite strongly toward the later view, and in the last couple years the CUR Quarterly has featured research on the learning gains from involvement in undergraduate research.

Monday, February 22, 2010

Shameless self promotion (cont.)

Mr. Forrest Hippensteel, an undergraduate researcher in my group was selected to present in a session where all the talks are invited. The Biological Fluorescence subgroup of the Biophysical Society wanted student speakers for their session, so they selected from among students who had submitted abstracts for the poster competition (these students also get recommendation letters from faculty advisors). On the basis of his abstract and my recommendation, he was chosen as one of the two presenters; the other student was a Ph.D. student. Interestingly, they didn't realize that he was an undergrad, which means that there was no special effort to select an undergrad for a broader impact or whatever--he beat the grad students fair and square!

I've had a chance to talk to a lot of postdocs and grad students at this meeting, and one point I try to make to them is that they can do good research at a PUI, and I point to my student's talk as an example. I've met a few postdocs here who I think will make great professors at PUIs, and one of them is working on a problem related to my research so I plan to stay in touch with her.

When should undergrads start interdisciplinary training?

The answer may seem obvious: ASAP! And I'm at a conference right now (Biophysical Society) that has lots of interdisciplinary stuff going on and I still realize that I lack a firm foundation in the bio side of biophysics, so that would seem like a further argument in favor of ASAP. And yet, I also see a lot of very successful people who didn't branch out until later in their careers, and the benefit was that first they learned how to do something very, very well, and really get that polish and depth, so they could bring it to other things that they do as well. I don't know that I can point to them and say "You know, think how much better they'd be if they'd just started doing interdisciplinary science right away!"

The easy answer is, of course, "It depends", and I won't claim that there is one magic answer. There's a place for fully cross-trained people, and there's a place for people who polished themselves by learning how to do something really, really well before branching out. The physics community makes good use of the later: People who got their Ph.D. in some sort of experimental work involving very, very high-precision measurement, and then took that depth of skill and applied it to other things. Often (but not always, of course) very high-precision measurements can only be done on systems that are in some sense simple or isolated, or that are amenable to detailed theoretical treatments so that the precise measurements can either be compared with theory or used to infer the strengths of various effects. These sorts of systems are often (though not always) more in the realm of traditional physics rather than interdisciplinary. Even when it is interdisciplinary, say, for instance, somebody making cantilevers for very high precision measurement of masses of molecules attached to them, the molecular system being studied may have all sorts of interdisciplinary significance but the techniques and mindsets of the measuring group will be more rooted in traditional physics.

I trhink that if we're training students for grad school, then encouraging them to focus for a while and become really good at one thing before branching out is a perfectly fine approach, at least for some students. It depends more on their mindset and inclinations than their raw talent; there are people who use abundant raw talent to branch out right away and effortlessly grasp a variety of different concepts, and there are people who use abundant raw talent to become really good at something in particular. However, if we're training people to enter the workforce after college, well, the workforce changes so quickly and is so uncertain that it's probably better to have some breadth so you can respond to whatever is out there.

Tuesday, February 16, 2010

What does a successful undergraduate researcher do afterwards?

So, let's say that you do a research project with an undergraduate. The student has fun, learns some useful skills, contributes something useful to the project, and maybe even becomes a co-author on a paper. So far so good.

And then that student gets a job in industry after graduation. And you have to write a report to a funding agency talking about the things you've done with students and how successful your students were. Was the project a success if the students didn't go to grad school? I'd say so, as long as the students did well on the project and learned something. The evidence for that learning might be a technical accomplishment, or piece of independent work, or it might be some sort of evidence based on techniques of educational psychology (e.g. the work of David Lopatto). Or it might be some other piece of evidence; the CUR Quarterly always has articles documenting benefits from undergraduate research.

However, we aren't talking about the general value of undergraduate research as a way of learning about science. We're talking about whether the funding agency's money was well spent on your program. If the project got scientifically significant results, and those results have been disseminated, that's certainly part of the evidence for success. However, funding agencies are also interested in training students. If a smart student goes to industry rather than grad school after doing undergraduate research, did we fail to retain some talent in the academic pipeline, or did we produce a well-trained member of the workforce?

I lean toward the second view; I see no need to clone myself by sending all of my best students to grad school. If I train people to go out into the workforce and do useful things (and make more money than a professor at a cash-strapped state school) I call that a win for the student. Now, if the student wants to go to grad school, great, but I would never try to sell somebody on it.

Do the funding agencies and professional societies agree with me? I'm not so sure on that. They talk about workforce development, but I see many discussions of the need to send more students to grad school. In fact, my inbox has at least one email on a program to get more students into grad school, and some of the externally funded undergraduate research programs on my campus are explicitly aimed at getting students to go to grad school. A good student who wants to acquire technical skills in a research lab before going out and getting a job is told "Thanks, but no thanks" if he/she applies for support through this program. (It may very well be that this student would be better off doing an internship off campus anyway, but if a smart student wants to do research with me for whatever reason, I'm not going to turn that student away on the basis of career plans.)

This would be fine if graduate programs were starved for students and the job market for Ph.D.'s had more positions than applicants. However, the reverse seems to be the case. To see the problem, go to any national conference, and in one room you'll have a speaker talking about getting more students to pursue graduate degrees (the audience may be faculty, or it may be a motivational talk for undergraduates). In another room, you'll have career advice for Ph.D. students entering a saturated and hyper-competitive job market.*

So, if a good student does undergraduate research and then goes into the workforce (and makes more money than a professor at a cash-strapped state school), is this a good outcome from the undergraduate research program or a bad outcome?

Maybe one issue here is that in some sense industry is a default: If a student doesn't go to grad school, where else will he or she go besides the private sector? So the student who does undergraduate research and goes into industry is in the same situation as the student who didn't do undergraduate research and then went into industry. How do you distinguish them? The work of Lopatto may be of some help, but in terms of success after graduation, well, tracking that student's salary and career satisfaction and technical accomplishments is hard. You won't know if your student is successful for some time, and you won't know how much of that success can be attributed to doing research. On the other hand, counting the number of students who go to Ph.D. programs is very easy: You just count. Done. Now write the progress report.

Anyway, what say the CUR members and other readers? Should grad school be the preferred outcome for students who do research? How can we measure the success of an undergraduate research program that doesn't send many students to grad school but does produce skilled people who go and get jobs?

*For extra irony, in one room you might have a speaker talking about the need to train our students to work in diverse, international environments to tackle the challenges of a global economy and global environmental, health, and technology problems. In another room, somebody will be lamenting the number of international students at our universities. I completely agree with the first speaker, and I hope that somebody introduces him or her to the second speaker, because I have a great idea for training students in an environment that has people from around the globe....

Wednesday, January 13, 2010

Long time no blog

And still not much time to blog, but I second everything said in this letter to Nature on interdisciplinary research for undergraduates.

I'll start blogging again...some day. I promise. Really. Honestly. This time I mean it.

Thursday, October 22, 2009

Undergrad researcher vs. technician?

Here's a question for anybody passionate about getting undergraduates into research: Is it OK to start freshmen off in roles that might be classified as "grunt" work?

I think we could all agree that a freshman shouldn't be making copies and fetching coffee for a professor. I think we could all agree that an undergrad shouldn't spend 4 years washing bottles (or, more likely in a physics lab, washing vacuum components, something I spent some time doing as an undergrad). And I think we could all agree that the ultimate goal of undergraduate involvement in research is for the student to be intellectually engaged and making intellectual contributions to the work.

But there are a few other things to consider: Everybody has to start somewhere, and the basic skills do need to be learned one way or another. A freshman might not have the background to take intellectual ownership of a project, but learning how to perform some routine tasks is a good starting point for somebody crossing the huge gulf between what you see in a typical teaching lab (I have on my desk a stack of reports written by freshmen who used tuning forks and PVC pipes and rulers to do an experiment) and what you see in a typical research lab.

The issue for me is not so much how a student starts but where the experience leads. Giving a freshman The Big Question that the group is trying to answer might excite the freshman and make all that bottle-washing and wiring and whatnot worthwhile, but giving an untested freshman a Big Question to take intellectual ownership of might not be so practical. Is it so bad to give the student some technical tasks as an introduction to the work and see what sorts of things the student does well, before assigning a question? One could envision a gradual process that begins with learning to run samples in the Whatchamacallit Machine. Eventually the student learns how to use the Whatchamacallit Machine and demonstrates that he/she can do it independently. Then you give the student a more challenging sample to run, something where some issues aren't clear, and now the student is trouble-shooting something and a participant in answering questions.

I think that involving students in the technical work of the lab before they're able to fully engage The Big Question of the research program is perfectly fine. I say this for 4 reasons:

1) Getting started earlier keeps people enthused and interested. If a freshman is excited about science, even washing bottles or whatever is better than work-study jobs in the cafeteria.

2) Frankly, we all have to learn the basic technical skills and the way a lab is run at some point. Why not give a Big Question to somebody who has some idea how a lab works and how the equipment works?

3) In anything but the most automated and tedious of technical work, there are still problems to solve and issues that require significant thought. As long as the student's brain is engaged, learning is happening and skills are being acquired.

4) Frankly, I think people too often choose scientific paths based solely on the question and not the work. The reality is that a working scientist does not spend most of his/her time on The Big Question. Theorists spend their time writing down models that don't work, reading papers, and writing code and trouble-shooting code and trouble-shooting still more code. Experimentalists spend a lot of their time building stuff and trouble-shooting it and fixing it and also reading papers. Observational astrophysicists spend most of their time in front of computers doing data reduction and analysis, not peering through eyepieces or pondering Life, The Universe, and Everything.

Every field has cool results that you can get to, but what do you want to do along the way? I used to do experiments, and I actually found certain parts of working with my hands enjoyable, but frankly it never appealed to me as much as analyzing data. At some point I realized that working with numbers and formulas was more fun to me than working with my hands, so I switched to theory. Conversely, I've seen students who found my theoretical projects cool because of the questions that I am working on, but they hated sitting in front of the computer and writing code. One of them then did an REU in experiment and loved working with her hands. Now she knows what she enjoys, and she looks for opportunities doing that. I call it a win for her.

So, maybe students shouldn't take intellectual ownership of a project until they actually know if they'll even enjoy the doing of the project, not just the end of the project. To that end, I consider it a useful educational experience for a freshman to start off in the grunt, technical parts of a research group's work, even if the student is not yet ready to take intellectual ownership of a question.

Eventually I'll blog about my own experience as an undergraduate in a research lab.

Friday, October 16, 2009

Promoting physics career paths

Via our fearless leader Chris Hughes, who has emailed all of the CUR Physics and Astronomy Division members, I learn that the American Physical Society is going to be posting a series of slide show presentations showing career paths in physics. Frankly, I think it's still a bit too heavy on the grad school/basic research side of things--most of our students will wind up in the private sector because basic research is over-crowded*, but it's still a good presentation. Hopefully future installments will say more about industrial careers. I'm at a school where it's hard to recruit physics majors when the engineering college is dangling employability in front of them. These videos will help, as will these numbers.

*One might ask why I'm so interested in getting students to do research if I think basic research career paths are over-crowded. The answer is that there are intellectual skills that simply cannot be taught if you aren't engaging with the unknown. When I lecture on things that I already understand, or even lead an interactive-engagement peer group inquiry-based buzzword whatever class, I'm in a different mode than when I'm working with a student to troubleshoot a new calculation that we haven't tried before. When we're dissecting a new article I explain things differently than when I demonstrate the solution to some problem that's been known since Newton's Principia.

Getting students into a mode where the answer isn't known, where they have to find it and solve technical problems along the way, and ultimately present it and defend their methods, is valuable preparation for a wide range of career paths. People in industry who are making a new product are, by definition, doing something new, and they'll have to troubleshoot things and solve problems that haven't been solved before and defend their work before supervisors and/or clients. So I'll defend undergraduate research in any and all arguments.

Thursday, August 6, 2009

"Everything We Teach Was Once Someone’s Research"

That's from the title of an article on the CUR website. There are a lot of insightful things in there on the history and future of higher education, and ideas from models of undergraduate research that may or may not work at your institution. Mostly, though, I just like it for the title (meaning no disrespect to the rest). Everything we teach was indeed somebody's research, and if we're not active in research we'll eventually be behind in the classroom. (This point is still controversial to some people at some undergraduate institutions.) Moreover, since everything we teach was once somebody's research, the research lab may be the best way to fully immerse a student in what it's really all about. There are things that I simply cannot do in a lecture, skills and approaches that I can only model when I'm working on something where I don't already know the answer. (Unfortunately, I'm a bit too timid to do that in the classroom, given the importance of evaluations.)

Friday, July 24, 2009

Advanced Physics Lab Classes

There's a post up at Confused at a Higher Level discussing some of the different purposes that one might have for an advanced lab class, and whether there's a sharp line between lab classes and research experience. Obviously a typical lab class cannot be an exact substitute for participation in a research group, but if you're looking for scalable models that introduce students to some of the important intellectual components of research, this is a topic worth pondering.