Wednesday, December 02, 2015

Advanced undergrad labs - survey

To my readers at universities:  I am interested in learning more about how other institutions do junior/senior level physics undergrad lab courses.  My impression is that there are roughly three approaches:

  • Self-guided or not to various degrees, students pick some set of predefined experiments that are presumably meant to teach pieces of physics while exposing the students to key components of modern research (more serious data acquisition; statistics+error analysis; sophisticated research instrumentation beyond what they would see in a first-year undergrad lab, such as lock-in amplifiers, high speed counters and vetoing, lasers, vacuum systems).  Sometimes students would work with an instructor to commission a new experiment rather than do one of the existing set.  This approach is what I saw as an undergrad - I remember running into a classmate late at night who had been doing some classic experiment confirming the \(1/r^{2}\) form of the Coulomb force law, and I remember three friends working as a team to commission a dye laser as part of such a project.
  • More topically narrow but intense/sophisticated labs.  For example, when I was a grad student I was a TA for a dedicated low temperature physics lab, where students chose from a list of experiments, designed some apparatus (!), had the parts machined by the shop (!!), and then actually assembled and ran their experiments over the course of a quarter.  It gave students a real sense of serious experimental research in its various phases, but only aimed to expose them to a comparatively narrow slice of modern physics.  I've heard of similar lab courses based on optics or atomic physics projects, and entire courses about electronics.
  • Some hybrid, where students do a combination of pre-fab experiments and then do a one-semester experimental project actually in an active research group, as part of their lab training and credit.
Questions for readers:  Am I leaving out some approach that you've experienced or run across?  If you're a faculty member in a university physics department, what does your department want/hope the undergraduates get out of these lab experiences?  Are there approaches to more advanced formal lab training that you particularly like and find successful (not counting having individual undergrads work in research groups, which we always encourage anyway)?  Students, were there particular labs or approaches that you really found valuable?


Anonymous said...

We (in a Western European country) had a combination of pre-defined experiments (with a little manual etc) that included questions on "you want to know A, how would you set up an experiment to do that?". So while the broad experiment you had to do was known, the specifics were up to you. It included the statistics, error analysis etc. Pretty standard.

In yr 3 we also had one "project" per semester (for about 2-3 weeks, each afternoon) where one would have to join a research group, and get immersed in actual research. In practice this meant walking with a grad student for a few days, and then getting an little experiment for yourself to do. (e.g. calibrate a pyrometer using a thermocouple, learning about thermal conductivity, thermocouples, optics, etc.)
One could choose the group - within constraints of logistics for the groups. The beauty of this was that you could do some reconaissance of research in general, and the type of research for groups you were potentially interested in for your MSc research project (with thesis) of 9-12 months in the 5th year.
(This was a requirement to graduate from college.)

I particularly liked this, as it really gave you a view of what hands-on research is, much more so than doing constrained experiments set up a number of years ago.
Doing a few of these projects exposed you to different research cultures of each group and subject.

Cesar Uliana said...

Hi Doug,

during my undergrad years in Brazil I had (in hindsight) a quite interesting experimental education. The labs courses ran through 3 years. The first one was run-of-the-mill statistics+erros analysis.

The second year was focused on two semesters with a small experiment to teach how to handle equipment, then a big one, usually 6-8 weeks long, with design, instrumentation, data collection and analysis. As an example, the first semester involved a "small accelerator". We had a cathode ray tube that need to be properly set up, then a capacitor and a coil to generate electric and magnetic fields to guide the electrons, and then a phosphorescent screen for detection. A typical week would include computer simulation of the capacitor to include finite size effects, then the next one would be a characterization to make sure the actual capacitor behaved as the simulated one. The point was not to teach accelerator physics, but to show all the parts of experimental practice. I've found that long experiments are more engaging, and require more understanding of experimental physics than short ones.

The third year would usually cover more elaborate equipment, so it had shorter experiments but with more advanced set ups, like vacuum tubes or crystallography.

Verle said...

at our university physics students must complete advanced lab from an option of either classic quantum experiment, astronomy observation, and or toilet lab option, which is most popular choice.

students must understand all physical concepts behind operation of toilet, including finite element based hydrodynamic simulations, chemical stain resistance, materials strength, and also full toilet dissection.

for final project students must design and build a new toilet in teams of 3-4 students. final day has live demo of toiilets, which must be fully functional for students to get full credit.

Professor Tico said...

We have in our department in addition to standard upper division lab course a physics of the supernatural component. This was originally optional but is now mandatory. Students must demonstrate proficiency with ghost hunting and identification of ghouls and spirits. Students gain valuable experience with specter-al analysis, capturing wraiths, demonic possession and exorcism, weather forecasting, astral projections, summoning the deceased, casting bones, and microsoft web products. Data have shown that this course ideally prepares students for real-life challenges they will face in the workplace in both academia and industry.

Dan Dougherty said...

We do a one semester "Advanced Experimental Physics" course where the first 6 weeks is a review of basic electronics, LabView, and data analysis. The rest of the semester is devoted to two different 4-week long "independent" projects. These can vary according to the tastes and motivation of the students from longer term "canned" experiments (e.g. Pulsed NMR) to real open-ended research projects.

Douglas Natelson said...

Thanks, Dan. I like the April 1 lab in particular :-)

Mary Kate said...

At my school we had a very standard third year lab course, where we all split into groups of two to four and worked on different pre-determined experiments. However, I think the most valuable part of the course was how the professor taught it. I often say that junior lab was the most influential class in my undergraduate coursework. While all my other courses taught me physics, lab really taught me how to think about physics. Our professor didn't let us get away with anything. If we chose a lab, we had to know it backwards and forwards. We had to understand the concepts, the equipment, and invent our own procedure. Our professor would often come around to interrogate us about what we were doing and if we didn't have a satisfactory answer about why we were doing it, he wouldn't let us move on until we figured it out.

I'm currently in my senior year, and each senior has to do a senior research project which applies what we learned in lab. This is supposed to be a year-long project, and most students work with one of the university professors or intern at one of the local industries (provided it is a physics-related job). At the end of the semester we have to write a report and defend our project in front of all our professors and peers-- with the lab professor infamously the most difficult to please (most students will go try to get him to reveal his questions beforehand, though he'll always hold the toughest one back to stump them on the spot).

Abhijit said...
This comment has been removed by the author.