Physicists Debate: Hardest Upper Division Subject to Teach

I wanted your thoughts on something. What, in your opinion, is the hardest upper division subject in physics to teach? Not necessarily the most difficult to learn, but to teach?

tiyusufaly said:

What, in your opinion, is the hardest upper division subject in physics to teach? Not necessarily the most difficult to learn, but to teach?

I think that this is subjective.

@George Jones
I voted Electricity and Magnetism and this was the text I used in that class:

tiyusufaly said:

I wanted your thoughts on something. What, in your opinion, is the hardest upper division subject in physics to teach? Not necessarily the most difficult to learn, but to teach?

I have only devised lesson plans and taught classes in EM including QM as it relates to radar science to graduate students and students close to graduating, primarily engineers.

In my opinion more depends on the students than the teacher and my hand-picked students, being motivated professionals, were a pleasure to teach. I learned something new every class I prepared and taught. I love EM and QM and like to think my enthusiasm was reciprocated by students and observers.

Textbooks were supplied by MIT originally supervised by Princeton professor Robert Dicke.

Teaching the associated labs on actual electronic devices probably formed the most difficult section. Students might readily grasp theory yet hesitate at manipulating actual EM fields. We solved this difficulty by employing experienced lab techs who, while indifferent lecturers, safely guided laboratory students through the nuances of applied physics.

If memory serves, beginning EM students attended three lecture days and one lab day per week. Advanced students attended daily morning lectures and afternoon labs with Q&A sessions before and after as new material was introduced. Frequent practical labs definitely helped students assimilate theory.

When a colleague is on sabbatical, taking over some of his/her classes for a year.

I voted for stat mech, although my real bête noire was classical thermo. All those partial derivative relationships, ugh!

I got stuck with it because neither of the other two people in my department wanted it or could fit it in. Eventually I found a textbook I really liked: Schroeder. However, I was able to use it only twice before I retired.

jtbell said:

I voted for stat mech, although my real bête noire was classical thermo. All those partial derivative relationships, ugh!

I got stuck with it because neither of the other two people in my department wanted it or could fit it in. Eventually I found a textbook I really liked: Schroeder. However, I was able to use it only twice before I retired.

As an undergraduate and as a grad student, I didn't like thermo or stat mech.
After having taught other undergrad courses in my tour as a VAP, I had an opportunity to teach stat mech... and I asked for it. I taught out of Schroeder... and learned to like it somewhat. I've taught it four times using Schroeder since then, learning more subtleties and appreciating more of the applications each time.

(Side note: I've been thinking about using differential forms in thermodynamics.. in particular, visualizing them. I presented a poster on the idea at an AAPT meeting. Hoping for some encouragement, Schroeder didn't seem to appreciate it... seemingly dismissing it as something that would be of more interest to a math person rather than a physics person. While true (at this time), I said that my approach for doing geometry on a PV-diagram is analogous to a Minkowski diagram for special relativity. That didn't change his response. I think I need to sell the idea better.)

I have taught all of the courses, and, for me, it was statistical mechanics. In a long teaching career at 7 universities, I have taught more than two dozen courses, but I had never taught statistical mechanics until the fall of 2020. I did teach second-year thermal physics from Schroeder in 2012.

4 weeks before the start of lectures In 2020, I was asked if I would teach fourth-year statistical mechanics as an emergency fill-in, and as a paid work overload. I had to prepare on-line lectures (COVID) for a course that I had never before taught, and that I had taken as a student 4 decades earlier.

As an undergraduate,
I learned an alternative name for statistical mechanics: "sadistical mechanics".

robphy said:

As an undergraduate,
I learned an alternative name for statistical mechanics: "sadistical mechanics".

statistical damnamics where i went to school

I have never taught any of the courses (though I would like to teach them all), but based in undergrad and grad experience with the topics, I would say that electrodynamics seems to be the more difficult one.

robphy said:

As an undergraduate,
I learned an alternative name for statistical mechanics: "sadistical mechanics".

Dr Transport said:

statistical damnamics where i went to school

For the classical stuff, I learned "thernomydamnics".

Well, I guess it depends on your audience. I voted for QM. Ordinary math/physics majors that do QM would be fine. Math majors that have done Hilbert Spaces would be a hoot. I can include Gleasons Theorem (modern POVM version) and show the basic mathematical foundation of QM is encoding the outcomes as the eigenvalues of an operator. The one that fills me with dread is philosophy students. For me, that would be murder. I would be saying all the time - who cares why it works - it just works. Remember, it is a model - like all of physics is. Do you want to go beyond that? Experience has shown that it is often a black hole you will never escape from and get nowhere. Then go over to the head of the philosophy department and 'suggest' they have a philosophy of QM course so that I can point such students in that direction.

Thanks
Bill

I have taught all four subjects on the list several times each. I voted for Classical Mechanics because that is the first course in the normal sequence of this list of courses. Because of that, few students have the necessary mathematical background in its entirety: three semesters of calculus, basic linear algebra and ODEs. Those who have the background have not seen it used. I will never forget a student in that class who came up to me and said, "how do I take the derivative of this expression where y is a function of x and x is a function of t?" "What do you mean?", I asked, "Just use the chain rule." He looked at me in astonishment and said "But that's math, it's not physics." He was a good student that just hadn't made the connection.

Klystron said:

I have only devised lesson plans and taught classes in EM including QM as it relates to radar science to graduate students and students close to graduating, primarily engineers.

In my opinion more depends on the students than the teacher and my hand-picked students, being motivated professionals, were a pleasure to teach. I learned something new every class I prepared and taught. I love EM and QM and like to think my enthusiasm was reciprocated by students and observers.

Textbooks were supplied by MIT originally supervised by Princeton professor Robert Dicke.

Teaching the associated labs on actual electronic devices probably formed the most difficult section. Students might readily grasp theory yet hesitate at manipulating actual EM fields. We solved this difficulty by employing experienced lab techs who, while indifferent lecturers, safely guided laboratory students through the nuances of applied physics.

If memory serves, beginning EM students attended three lecture days and one lab day per week. Advanced students attended daily morning lectures and afternoon labs with Q&A sessions before and after as new material was introduced. Frequent practical labs definitely helped students assimilate theory.

I agree that it depends very much on the students. As for me, the most difficult thing is to apply knowledge in practice, because theory is just to remember information and how the lecturer delivers it.