Print This Issue

Talk About Teaching and Learning

February 15, 2011, Volume 57, No. 22

Teaching the Next Generation

Mark Devlin

When working with the undergraduates in my lab I often reminisce about my days working in the X-ray astrophysics lab at the University of Wisconsin. There I spent three years building electronics and cryogenic equipment for my advisor. I worked late nights and weekends often side-by-side with my advisor as well as on my own. The experience and skills honed in that lab laid the basis for my abilities as an experimental astrophysicist. I used them in graduate school and I use them as a professor at Penn.

There is no question that having experience doing research can be an extremely valuable, if not essential, part of an undergraduate education.  Almost all of an undergraduate’s exposure to faculty is in a classroom setting. The student might feel they are passionate about a certain field, but how are they to know what doing original research is all about? Providing them with real research experience prepares them for the next step in their careers whether it is graduate school or some other advanced training. It also gives them a chance to decide if the direction they have “chosen” is right for them without taking too big a leap. Finally, they become much more attractive candidates for the best graduate schools and also to future employers who crave people with practical experience.

While part of my motivation for involving students in my research can be traced to my undergraduate days, I had more practical reasons when I first arrived at Penn 14 years ago. At the time there was no real astrophysics program. We had no reputation and hence we had no graduate students applying to the program. In fact, I would not get my first student until almost four years after I started. My first experiment was very instrumentation-intensive, requiring us to design and build many components. The only way to get it all done was to hire an army of undergraduate researchers.

With five or so undergraduates we built an entire telescope pointing system and the software to run it. Working with colleagues at Princeton we took our Cosmic Microwave Background experiment to the mountains of northern Chile in 1998.  While the students did not accompany me to Chile as they do these days, they were an instrumental part of our success. The data from that experiment was phenomenal.  It got us a front page article in the New York Times, and, more importantly, it got me tenure at Penn.  So, I can say from personal experience, when done correctly, hiring undergraduates benefits the students, the faculty, and the School as a whole.  

So, the question is, how to involve undergraduates in your research and have it be a positive experience for everyone? I have had about 50 undergraduates work for me over the years and have developed a few techniques that seem to work. 

The first thing to understand is that there is a temptation to feel that the students should be happy for the experience and that pay is a secondary issue. This will only work in the rarest of cases, so you will need to have some resources to pay them. I got one of my best students after he wrote a program to scan the job listings every day optimized to his interests and the amount of pay!

Most federal grants look very favorably on projects that involve undergraduates. I put a small amount for undergraduate research in every grant proposal I write.  A typical student will cost you about $3,000 over the course of the year if they work during the summer.  I post my jobs on the Penn Student Employment Office web site (www.sfs.upenn.edu/seo/).  I keep the jobs posted all the time just in case a “must have” student comes knocking on my door.

When you interview students try to separate the students who are simply looking for a paycheck from those who are really interested in your research. Of course, there are those who are simply looking for a paycheck but are phenomenally good at something (like that programming student!). 

Besides the boilerplate questions about skills and knowledge, I tell them one very important thing: I am hiring them because I need a job done. Now, don’t get me wrong. They will learn a huge amount by working with the more experienced members of my group. But I am entrusting them with tasks that need to get done for my research which is often on a tight schedule. If they start something and then lose interest or just get too busy, I am left holding the ball with a half-finished project. Too many of those will put you off hiring undergraduates forever!  I believe that this serious and honest approach increases their ownership and commitment to the work.

I used to think that I should provide a separate meaningful project to each student. This technique has the inherent problem that the weaker students will get completely lost, require too much of your time (or worse, the really productive members of your group), and eventually fail. A much better model is to group students together on a single, more substantial project. This way you can meet with them every few days. Between meetings they can help and learn from each other. Right now I have four students working at building a new cryogenic test system. When I walk into the room I can see them working, interacting, learning, and getting the job done! Watching the students actually produce something, some for the first time in their lives, is a true joy. This is probably the main reason I keep hiring so many each year.

I tend to hire my students for the summer after their freshman or sophomore years. My expectation is that they will come up to speed in the lab or with some research topic over the summer. During the school year they might spend ten hours a week making reasonable progress. However, over the next summer, the ones who continue should be fully productive members of our research team. They will be able to work with the younger students easing the burden on my graduate students and postdocs. The process is not exactly self-perpetuating, but by their senior summer, a few of my undergrads have been more valuable to the lab than some second year graduate students.

A side benefit from having so many undergraduates working for you is the great advertising you get. My students have gone to graduate schools all over the country. I encourage them to go to places that have a positive research environment. If everything goes full circle, they will be interacting with students at their new schools.  If they had a great research experience at Penn, they are sure to pass along the word. One of my finest graduate students actually came from the lab where I did my undergraduate research.

Of course my experience is restricted to working in a large group with many resources and experienced people to help with the “undergraduate horde” that I hire each summer. Consequently, some may think, “This can’t possibly work for me.” Really, it can work. We can all teach; that is why we are here at Penn.  If you can take only one student on, give it a try!  Take a minute and write down a list of ten fun and short projects that you wish you had time to do.  I would bet that you can find a student who will be a good match for at least one of them. What is the worst that can happen? I assure you, your successes will provide enough incentive to balance any challenges you might be imagining. Good luck!


Mark Devlin is the Reese W. Flower Professor of Astronomy and Astrophysics in SAS and
the 2010 recipient of Dean’s Award for Mentorship of Undergraduate Research.


This essay continues the series that began in the fall of 1994 as the joint creation of the
College of Arts and Sciences and the Lindback Society for Distinguished Teaching.
See www.upenn.edu/almanac/teach/teachall.html for the previous essays.


Almanac - February 15, 2011, Volume 57, No. 22