TAing

So I got to Madison last weekend, and we spent this week doing TA training. I thought this was a good idea, since I need all the training I can get to deal with 3 sections of 24 kids each. The training was pretty good quality too, except they kept leaving gaps which our profs were to fill in (how they wanted the labs done, what we should go over in discussion section, what to do for weekly quizzes, etc.), but this didn't really bother me. The profs would fill in these gaps, right?

Then the other 9 TAs for my class and I met with our two profs. Pretty much all they said about labs and discussion section was, "grade to an average of 80 with as wide a distribution as you can." No attempt to standardize grading across TAs. Little input on how or what to grade in labs or how to make the quizzes. Overall, not much guidance (there was certainly some, i.e. "we'll focus on conceptual issues in class, so you should focus on working examples in discussion section," but that's kind of vague). Thankfully, a couple of the other TAs had worked on this class before, so we got together afterwards and sorted a couple of things out, but it's still pretty much every TA for himself. And it's not like the profs seemed uncaring about the class -- they seemed like nice folks and their reviews seem to agree. I'm sure that I'll get things figured out after a week or two, but I was expecting a little more guidance.

Other TAs, what has your experience been?

Thank god there are people at Berkeley who care about pedagogy, even if those who are most dedicated are often senior graduate students. It's been a bit mixed, but in general the department has worked out very clear instructions on what TAs should do for lower division courses, so my job is straightforward. We also had a university wide day long training. Grading is very carefully standardized, although I haven't done any yet. The head TA for large courses helps to keep us on track.

It's good to see that some places do it in a more organized fashion.

No one else is TAing eh? Or maybe they're all hard at work...

Hey, I had (am having) a similar experience, except I think mine was even more loose and unstructured. There was general graduate student orientation which had a few sections on effective teaching. It was by and large helpful, but very general. I think it was a Philosophy professor that spoke to us about communicating, teaching, and engaging students.

As for the actual lab instructor, we met twice as a group with the other TAs, and both times lacked details. I tried to ask specific questions about grading criteria and in lab teaching expectations (formal lecture or just go-at-it and answer questions later???), but only got a vague answer about "grade to a 88%" and "encourage them to write lots". Other incoming TAs were asking me questions, so I think that they felt even less guided then I. Many were international students who had only a vague idea of what expectations were going to be for them. I can't imagine what it would be like to have some language difficulty, be in a foreign culture, and then be expected to lead and direct 2 lab sections.

Fortunately I only have one section (a small evening lab for second year students) but the first week I filled in for another TA who is held up with VISA issues. I came up with a simple grading system (start with 8.8/10, add these points for this, subtract these points for that...) so I could at least try and be consistent.

It feels quite informal from the instructor side of things, but the undergraduate students want to know more details (which I can totally understand). So I feel like I'm creating structure on the fly where very little actually exists.

Even though I would have liked more guidance with the grading standards, I feel very good about how the first 2 weeks went as a TA. The quality of the student's work has been pretty good, and they seem to understand me pretty well (or maybe not and they're getting it despite me!). Maybe the best way to learn teaching is just by doing....

Good luck to the other TAs,

Marten

When I was a TA there was a lengthy "training" session for TAs of all majors. However, the focus of the lengthy training session was just a bunch of politically correct stuff I had heard many times before.

Unfortunately, I believe with most schools there is very little training given to help teaching assistants teach. It is like "here is a lab manual and coordinate with the other TAs to determine who takes what time slot".

If you want to be a good TA, then I would recommend you get some advice from veteran TAs who are both good at being a TA and who also enjoy teaching. These people not only like teaching, but they also like teaching people how to teach. Maybe get some of the other TAs in your class together to chip in and treat these veteran TAs to lunch for some meaningful TA training. Maybe summarize what you learn from them here on physicsgre.com Actually, if you contact me and agree to write a summary article of the TA tips you learned from these lunch sessions then physicsgre.com will pick up the lunch tab for the veteran TAs. I'm being serious so please use the contact form or send me a private message if interested in getting some money (and t-shirts) to treat veteran TAs to lunch. It is OK if it is specific to your class as this information will still benefit other TAs.

I do have one tip/comment on grading. I think it is important to try and have some consistency with grading with the TAs in your class. For example, depending on your definition of fairness, you probably don't want to be the TA who gives most people C's when the other TAs in your same class are handing out A's and B's like candy. I would coordinate with other TAs and come to an average grade consensus as the official-department-party-line-average-grade theory may not necessarily agree with experiment.

Also, in case I was giving the wrong impression, most of the credit for my (relatively) good and guided teaching experience should go to the experienced head TA leading my course. Support has been much less even from the department and professors.

Edit: removed details

I'm curious, have any of your training sessions addressed the issue of how strict you should be on students for accuracy in the labs?

From what I've heard, and from my own experiences, undergrad physics labs are notoriously dysfunctional, hence the joke:

"You enter a class and see a lab experiment set up...
What subject is it for?
If it wiggles, it's biology.
If it blows up, it's chemistry.
If it doesn't work, it's physics."

So the question is... do you take off credit for honest data that isn't close to the accepted value, which only encourages students to make up the data... or do you focus more on effort and a sound analysis. I once had a TA give us a 7/10 for a lab that we kept getting 50% error for even though we followed the procedure as closely as possible, and the TA tried it and couldn't do any better... needless to say, from that day on, all of our labs had data within 5% ...

Another issue is, do you make them write tedious formal lab reports, that include rewriting the procedure, theoretical background, equipment etc... which are already given in the lab handout? I had a TA that made us do that... my lab reports that semester took a whole day to finish and were upto 15 pages in length. Other TAs only asked for data and analysis, and were thus well liked lol. I can understand making the class write one formal lab report per semester to make sure they can do it, but the rest should be informal.

At least I won't have to torture students... at PPPL they don't make us TA, probably because the lab is off campus and doing so would be logistically impractical... but if I'm ever in charge of a class one day I will try to maximize the educational value of the labs and minimize the pain.

Although I haven't begun to TA, (the school year hasn't even begun yet) I do have a couple opinions on how grading should be done. Even if the undergrads come away with no knowledge of physics, I believe it is important to have them understand the process in which science is conducted. To have them come away thinking that experimentation is conducted by fudging their data is unacceptable. They should not be graded on the quality of their data but on how well they conducted the procedure(did they follow directions?, were they careless?, ect.). The majority of their points should come from their analysis. Did their data support the hypothesis in the lab notebook? If so, how? What were the major sources of error(and don't say human error!) and how did it affect the data? I am also a sucker for neat and organized work. I, under no circumstances, accept work done messily with regular, wooden number two pencils. The combination of crappy handwriting, smudges due to them trying to erase their extremely dark handwriting grosses me out. Those that have naturally crappy handwriting(they always seem to use number two pencils), I have them use mechanical pencils with hard graphite. For those that say it is unavoidable, I believe, are wrong. I have witnessed these people write legibly because they slowed down and carefully wrote their letters. As for having to write down the procedure and such, I don't think they should. Even if it was meant to make them read it, it doesn't work. For all except those that are truely interested, it is a time sucker and their more important work suffers because they are hurrying to complete it and they become careless. Anyway, those are my "couple" opinions. Good luck in creating a system that works for you guys/gals! The lack of proper direction must be discouraging.

My high school physics teacher was very fond of that saying about labs.



In my case, I never even look at the lab notebooks (this is a non calc mechanics class). We give a quiz every week on the previous lab, and the quiz is open lab notebook. So I couldn't really grade them on accuracy if I wanted to (which I don't). Instead, there get quizzed on procedures, conceptual issues, etc.

I've been encouraging them to use the notebooks as much as possible -- we'll see how much they do as the term goes on.

I also forgot to mention I have three days(8am-5pm) of TA training with a little department introduction thrown in. It seems, and I am assuming because of the length spent in training, that Davis is pretty thorough in how we should TA. If it turns out that I am spending nine hours listening to non-sense then I will be pissed. I like teaching though and hopefully I can get across the necessities, command respect and be able to grade consistently while being an approachable, nice TA. I, however, have moments of extreme annoyance when people don't put in a decent amount of effort and as a result get in a bad mood and everyone pays. Because of this when I grade homework, I organize the pile into "give a ***" and "don't give a ***" and grade the latter pile last so that I am sleepy and get pissed off easily. The only trouble is that I can get in a really, really bad mood and end up making comments that, although not outright disrespectful, are round-about ways of telling the student "you're being stupid here." I think everyone knows how I feel(even professors) so I am sure I am not alone. A funny story is that I actually got politely fired(the professor said that since I was taking such a full load that I wouldn't have time to grade) from a calculus-based mechanics course, hired for a calculus-based E&M course grading lab manuals and subsequently made to ease up and then got politely fired again, got hired for another calculus-based mechanics course from a different professor which had frequent talks about my grading(he said I was being too thorough) and then got hired the next year by my first professor since the last one went on sabbatical. I seem to always expect too much but I don't see it. If the writing is terrible, mark them down; if they fail to show their work, mark them down; if they forget to draw pictures, mark them down; copy another's work earns them and the other person a zero. But oh well, back to studying for qualifiers.

Pedagogically speaking, marking students down for failings that do not reflect a lack of understanding is probably not a good idea. Much more important than enforcing some set of principled grading "standards" is motivating students and encouraging them to learn. Grades serve an effective tool when they encourage people to be diligent students and learn the materials. Sometimes that can mean ensuring that they've figured out how to solve every problem correctly, but the value is in practicing problem solving, not having solutions. Severely marking down people who make good faith efforts may likely contribute to turning them off of physics all together. I think exams are probably a better place to differentiate success.

The other virtue of grading homework easily, of course, is that it's a lot easier to manage as a grader.

Update: one of the profs for the course I'm teaching seems to care -- he's been showing up to the weekly TA meeting and giving a lot of helpful feedback on how to teach. I've yet to meet the other lecturer four weeks into teaching his class.

Wow Grant, that's a generous offer. I'll have to see if I can manufacture the opportunity to take you up on that. Sure could be helpful, I think, to new students.

Marten

Well, one term TAing down, and (hopefully) one to go. How did people fare?

The first few weeks were kind of rough. During TA training, we had been given a short presentation on how to do group work (including this wildlife documentary like video and several papers claiming it was the superior way to lead a discussion section -- as opposed to just working problems on the board). So I tried a mix of lecture review, board work, and group work. This worked well for a while, but the pace of the lectures started to pick up, and it became clear that I didn't have time to do all three. So I requested feedback from the students about which parts they liked. To my surprise, the average ranking from most to least useful was board work, lecture review, and group work (my ranking was group work, board work, lecture review).

I had already decided to cut back on the lecture review, and mix it in with the board work, but the big questions was what to do with the group work. Research showed that it was effective when done correctly, and I was following the recommended format fairly closely. Who was wrong? My current theory is that board work can give a false sense of confidence -- it's easy to see a problem done and then think, "I can do that. No problem." But it's not always true -- working the problem through is the way to find out. Combine this belief that board work is as effective as group work and the fact that it takes less time to do a question on the board than in a group, and the perceived efficiency of board problems could explain the reasoning behind their ranking. They also had plenty of opportunity to do group work on the homework (and many of them did), so perhaps discussion section group work was superfluous.

In the end, I cut the group work, both because it was unpopular and because the increased class pace made it too hard to both do group work and board work (I felt some board work was necessary). My students seemed to do fine, but I still wonder if that was the right decision. At any rate, my evaluations were good, so I managed to please them (although everyone hate my quizzes... I was only starting to get the hang of making them at the end).

Anyhow, this term is completely different. I've been assigned to the same class, but with a prof who does things differently. He and the head TA have pretty much planned out everything -- the discussion sections will be pretty much all group work and they have made up worksheets for all the TAs to use. Quizzes are taken care of in a similar fashion. They seem to know what they are doing, and it certainly means less work for me, so that is nice. On the other hand, I had a bunch of ideas for how to do things differently this term, and now I won't get to try them. I'm curious to see how this setup works out.

Physics education research shows that group work is MUCH more effective for learning than lecturing. It's funny but true that often students don't know how they learn best.

Here was from my favorite TA evaluation:

The truth:

My evaluations told me exactly what I already knew, which is that I really should have spent more time preparing so that I was better organized and could have presented the material more coherently. Intro mechanics/thermo for premed types turned out to be much harder than I expected, and for some parts I struggled to find a good way to explain things.

But no teaching this semester. I found a theorist who's flush with money so I'm doing research instead! (I think I'm the only first year interested in theory who's getting funded for research.)

Seriously, who is doing this research? How do you quantify better? I don't believe group work does any good.

Group work doesn't work because groups will almost always have a student who wants a free ride. I have never worked well in groups, nor have I ever observed other groups of students doing any effective work. Listen close to the groups and you'll hear things like, "this blows" and "if they're not going to teach they should just let us go."

Students are paying for an education. Please don't waste it by making them teach themselves.

zxcv, I thought you were a girl.

Hey dlenmn! That post was very helpful for me. I'm applying to UWisc and I'll probably have to TA there. I have some questions which may sound like a joke, but I'm a non-native English speaker, so I'm not sure if I'm getting some meanings right to better understand how you organized your classes. Correct me if I'm wrong:
Lecture review: it's a revision of the contents given by the prof during lectures, no problem solving at all.
Board work: This could be either you solving problems in the board or getting some of the students to work out problems in the blackboard. It's probably the first one only, right?
Group work: got it.
Thanks!

Yeah, and by that metric, doing problems on the board doesn't work because some students will zone/pass out...

Sounds like you got it. I'd say "summary" rather than "revision" for lecture review. As you suspected, board work is just the TA writing on the board (although some TAs have students write up their groupwork on the board if they're done early).

Are you thinking of doing theory or experiment? If it's theory (meaning you're going to spend many terms TAing) you might want to consider taking a delta program class early on. I hear they're not that much work, but they do a good job teaching you how to TA (group work and the likes seems to be becoming more common in physics classes, so perhaps it's best to learn how it's done). They're not worth so much if you do experiment, since then you're only going to be TAing for a term to two anyway.

Board work works because the students who don't want to pay attention aren't a distraction and impediment to those who do.

Thanks for the tips dlenmn! I will read through the stuff you mentioned. I'll be doing experiment (I hope) in IceCUBE. I've been TAing during this last year and I've tried several things with varying success.
What it seemed to work for me was to do a very light summary first of the concepts, I'd then start with the problems that are conceptually important to stress some material. Then I would leave then do their work in groups (3 people per group maximum), giving a hand if somebody got stuck. At the same time I would ask them questions about how they solved the problem and why it can not be done in any other way without contradicting what they've learned. I'd define some loose time span to do an specific problem and then I'll do it in the blackboard eliciting participation from them (and also trying to trick them in some steps to see if they can figure it out, yes I'm a good boy.)
Are you on theory?

Encourage group work only at the detriment of learning. I think it would be better to go through a problem step-by-step illustrating the places where you are applying concepts discussed in class.

Group work is not allowed on tests.

By the way, I think most students will give good reviews if you are not a complete asshole, you don't take up too much of their time, and you tell them what to expect on tests.

Although I agree with some of twistor points (I mean, groups will probably contain slackers that are there for the free ride only) there are also bright points in group work learning.
During this past year I found that some students won't ask me questions while working alone (I'm talking about freshmen here) maybe because they thought that the question was too simple or stupid to be asked. While working in groups, when they can't figure something out they are more likely to ask you things because, at least, there's somebody else with the same doubt, so the question may not be trivial.
My problem was always to keep the pace while working with groups. You can't wait until everybody is done to make sure they did the problem, I sometimes felt guilty to stop at a point and start doing the exercise on the blackboard while some students were still fighting with the problem but in any other case it would be making the class go at the speed of the dumbest kid, which is not a great idea.
After the students got used to the rhythm they were able to work nicely, share ideas with their partners, feel confident to ask me in case of trouble and, the most important thing, to develop their own intuition on how to solve physics problems. When I was doing board work only and they watch me doing the problems they said "it's a piece of cake", but when they were alone they didn't even know how to get started.
I'm sure there is no final word on this, it depends on the person teaching and the persons that are learning and different classes could have different dynamics, but I think there are positive things in group work.

PS: When I talk about groups I mean 3 members as maximum, not a roman senate

I would agree with this completely. If you are getting awful reviews it is not a coincidence and you should do some analysis of yourself as a person.

Here's an example I have handy:
http://dx.doi.org/10.1119/1.2162549

Sorry to disappoint! I do care about gender equality issues, though, so perhaps that's why you were confused.

Colorado tried to replicate the Harvard model with mixed results: http://prst-per.aps.org/pdf/PRSTPER/v3/i1/e010107. They observed no gender gap reduction, although most students improved conceptually overall (note also that improved performance on a conceptual test doesn't necessarily equate to improved problem solving).

My personal opinion is that group work blows, and those who want to do it can always do it on their own time. (And I'm a girl - I'm supposed to be "collaborative" and all that BS). Board work is a wonderful thing, especially if students don't see worked examples during lecture, because it helps provide a solid model for problem solving structure. Many students in intro classes have trouble extracting relevant information from the problem and creating a "mental map" to the solution; they tend to just throw numbers at equations and jumble them around until something comes out. Methodical board work can clarify the process by which one should approach a problem.

The other advantage to eliminating group work is that it takes a lot of unfair pressure off the perceived "better" students in the class. Some of us, who may or may not necessarily be better at physics, nevertheless are very quick to figure out how to approach a new problem type; other students immediately come to us and expect us to teach them physics, which we can't do because we're still learning too and we honestly have no clue how we decided to do what we did. So we get ourselves confused trying to explain things, they get completely lost in our incoherent rambling, and nobody wins. It's much better to have somebody who actually knows what they're talking about do the teaching.

I was initially skeptical of group work for some of the reasons mentioned here -- my group work experience from before college was fairly negative. However, when I tried it in my sections last term, it went smoothly. I think a lot of it has to do with having the teacher interact. In my previous experiences, the teacher would give group work and then sit at their desk and do whatever it is that teachers do in such situations. This lead to all the aforementioned problems.

By continuously walking around and interacting with groups (even if they don't request it), you can mitigate some of those problems. If it looks like a group is spinning their wheels, but not asking for help, you go and see what they're up to. If it's clear that someone isn't pulling their weight, you go over and ask the person about the problem they're on. Simply being among the students can keep them more on task. As you get to know the students (and them you), the atmosphere became pretty relaxed. Interaction didn't solve all the problems with group work, but it certainly made it work better (to the point where I thought it was very useful).

That said, I don't think it makes sense to do only group work -- I think it makes sense to do it along with board work. Like other have said, seeing worked examples (with step by step guides to solving similar problems) seems like an obvious thing to do.

Anyhow, I see this term's layout as a worth a try. I'd like to see the results. They've done this setup before, and it did not cause the students to spontaneously combust. They did fine, so some of the more dire predictions for group work probably aren't true. Talk and opinions are cheap. I'll go with the plan and see how things turns out.

What setups(s) have people seen board work done with and what were the results? (Group size, teacher interaction, mixed with board work or straight, etc.)

Related thought: Education research is hard to do in part because there are many little things that factor in to the end result, but are often hard to first identify and then replicate elsewhere (Feynman had a bit on that I think). Hopefully they'll sort things out eventually, and we'll someday conclusively know what the results of different methods are.