Making a classroom discussion an actual discussion

by Harry on September 12, 2016

In this post I mentioned a time that I had my small (21 person) discussion based class recorded, and then watched the video with several colleagues (and 3 students I invited who were actually in the class). Someone observed, pretty quickly, that the discussion had a kind of ping-pong feel. The students were all willing to talk (event the student who told me in the previous class that she was ok with being recorded as long as she didn’t have to speak in the discussion), but they were all just talking to me. We were in a circle, so it was entirely possible for them to talk to the whole class, but something I was doing was preventing that, and doing it, anyway, was not what they were used to (all but one were first-semester freshmen). What I was doing, specifically, was affirming, or rephrasing, or gently correcting, or responding to, what each of them said, preventing a flow of conversation. And, of course, responding to interesting things each one said, with something else interesting for the whole class. So, it wasn’t wholly bad, and clearly my motives were good. But it was a failure, something like 21 separate and not that great tutorials, all happening at the same time – and I would say it was a fair representation of my classes up to that point.

So, how to change that? One commenter said “I would love to hear, either in the comments here or in a separate post, what strategies you’ve developed to get past (or to some extent deal with) this problem.” I held off partly because it was summer, but mainly because I wanted to wait till I had, as it were, watched myself in action, to see what I do now that makes class discussions real, full on, discussions, in which students are giving one another reasons, listening to one another (not looking for my approval) and improving as thinkers and talkers. So, the semester has started again and, luckily, I am teaching two smallish classes (one has 26 students, juniors and seniors; the other has 22 freshman).

What did I notice?

I am relieved to observe (and have this confirmed, in one class, by a students I have recruited to observe and critique my teaching) that, indeed, the class discussions have flowed well, with everyone addressing their comments to the rest of the class, and everyone else seeming to listen; with most comments being responses to comments that have been made.

One thing I noticed is that I am very explicit in telling the class this is what is going to happen – I make clear the expectation that they address their classmates, not me, even though I will usually be the person asking the questions and setting the agenda, and moving things along.

I am also very deliberate about the questions I ask. Normally, class discussions are a response to some problem or prompt that I have devised, which related to a problem that arises in the reading, or directly about the reading itself, and which they have written down either on a handout or (if its short, and this is generally better) on a ppt slide. So the questions are usually very specific but, obviously, questions about which reasonable people can disagree – and about which I can expect disagreement in the class – and which require the giving of reasons.

The main thing is this. I need something to inhibit me from constantly jumping in with either i) interrupting reassurance or ii) some interesting, pedagogically valuable, comment (of which many come to mind). So I’ve engineered some sort of gestalt switch in my head. When a student says something, instead of thinking that I am depriving her by not responding (either assuringly or interestingly) I think to myself that I am depriving her precisely by responding – depriving her of the interaction with her peers, the reasons they can give to her, and the opportunity to surprise them. If the conversation ebbs, or if some particular strand is, in my opinion, played out, I step in an prompt the discussion with further questions, and often with low-pressure cold-calling (I also deploy gentle cold calling as the discussion moves along – if someone hasn’t spoken yet, or recently, they obviously go to the top of the queue, but also if someone looks like they are thinking, and haven’t yet spoken, I’ll call on them to see if they have something to say). My rule of thumb is that on average at least 4 people should speak in between every time I say something substantive (as opposed to just calling on another student), and as long as I keep to that, discussion goes well.

Two observations. I have had to learn to resist the incredibly powerful pull of needing to present a neat and tidy package to them, so that all issues are clearly articulated and summed up, and no ends are left loose. I do some summary at the end of a discussion, and always explain the point of the discussion and say what I want them to have learned (except when I want them to be kept in suspense until the next class period). But I try to keep it till late in or the end of the discussion, and I do not (usually) summarize by giving them the full correct truth (Brighouse edition). This is the most difficult thing for me – all of my training and instincts are to tell, not show, and they are good instincts for some purposes but definitely not for leading a discussion. Also: this is made complicated by the fact that sometimes you absolutely should intervene, because sometimes a student says something that really mustn’t get lost in the discussion but will get lost (and its significance thus not noticed by the other students or even by the student who said it) unless you emphasize it. So, sometimes you have to intervene, but much less often than you want to.

This is all so much more mentally taxing than carrying on 21 separate conversations one after another. You have to follow along the whole conversation, keep everything on track, interpret what the students are saying, remember what needs to be highlighted at the end…. and be sensitive to the needs of each student (some of whom need more drawing out, others more reining in, etc). And especially difficult with freshmen who are, typically, more inclined to be deferential and, just because they know less and have less experience, have less to say. But, it does seem to be better pedagogy, and my guess is (though where’s the rigorous evidence?) that more of the learning I want to happen actually happens.

Anyway, that’s my best effort at saying what I do that seems to work better, at least, than what I used to do! There’s nothing magical or remarkable about it, and I learned to do what I do largely by watching other teachers who are better skilled (in particular, 2 of our grad students, and also my now colleague at the Center for Ethics and Education, Paula McAvoy). I wouldn’t even bother writing about except that i) I was asked and ii) I know from talking to students that this is not pervasive practice. Please comment, offer improvements, suggest alternatives, share experiences, etc.



Kevin McDonough 09.12.16 at 12:53 pm

This is incredibly useful. Thanks, Harry. May have further questions or comments after I teach tomorrow evening!


M Caswell 09.12.16 at 2:05 pm

Great observations. I use a similar technique to help myself, psychologically, remain quiet while students talk something out for themselves- I tell myself that by remaining quiet at just the right times, I am deploying my pedagogical expertise, such as I have any.

You mentioned calling on students, and even a ‘queue.’ What would happen if you never called on anyone at all? 21 students might be right around the maximum where that approach could work.

A couple of other thoughts:

1) Students have to know each others’ names well enough to track the conversation, for example, by referring back to a point made by Ms. So-and-so 15 minutes ago. Collegial use of names promotes the reciprocity and recognition a healthy conversation needs.

2) In order to prevent becoming the center of the discussion, as the one through whom each remark must be mediated, I think it is extremely helpful to begin with a question that is really a genuine question for me- that is, a question I do not know the answer to. Once students see this, and if I’ve picked a question they also find compelling, the conversation becomes a cooperative inquiry, rather than something administered by me.

3) For real conversation to take place, some floundering must be permitted. The leader has to somehow distinguish between the floundering that is part of learning, and the floundering that is truly a waste of time. Not easy.


Phil 09.12.16 at 3:22 pm

Next question: how do you do all of the above in a classroom with fixed rows of seats? I’m currently working on not presenting myself as the centre of attention and the fount of all knowledge, but classrooms where everyone is forced to look at me, rather than one another, really don’t help.


jdkbrown 09.12.16 at 3:54 pm

Harry: I just wanted to leave a note of deep appreciation for your teaching posts. They’re one of the major touchstones for examining my own teaching.


passer-by 09.12.16 at 4:46 pm

As a long-time lurker, I second jdkbrown: those teaching posts are really great. I’m really looking forward to the discussion on class discussions, because I really struggle with it (as a young assistant prof in Europe).
It’s particularly true for me because I hated them as a student. All the profs I remember best were mostly exceptional lecturers. The European country where I did most of my studies emphasized more traditional teaching methods. The first time I came accross real discussion classes was as during the year I spent as a grad student in the US, and I was extremely frustrated with the discussions. I wanted to hear from the teachers, not from other students during class. Small, graduate only seminars could lead to some excellent discussions, but even then, the times I felt I had really learned something during the discussion were rare. I dropped a class with an anthropology prof whose work I really admire. His extremely popular seminar on violence had 25 students, including several seniors, and he was all for discussion. During the second week, we spent two hours “discussing” Marx’s 18th Brumaire, and all the intelligence of the students could not compensate for the fact that as far as I could tell, almost none of them had any prior knowledge of Marx, Marxism, 19th century France or even Europe, which kept the “discussion” at a very superficial level, in spite of how smart the students sounded. I dropped the class when I concluded that if I wanted to discuss African independance wars or paramilitary violence in Latin America with people who knew no more or way less about the topic than I did, I could do it anytime (and actually did do it constantly – grad students talk to each other a lot outside class, after all).
To a great extent, after all, I feel the same as a researcher – discussions with colleagues are obviously an essential part of the way I advance my research and develop intellectually, but if I go to a conference or a seminar, I really want to hear what the speaker has to say, in a structured fashion, with results and all, not have him lead us in a discussion of the material he has gathered.

I do recognize that my preference for much more traditional, content-heavy lectures would only fit a very small subsection of the best students. But it makes it very difficult to me to really imagine how a successful discussion should look like, and how to achieve that outcome…


Human 09.12.16 at 4:58 pm

I’m interested in techniques you have found successful for reining in the too-talky ones!


kidneystones 09.12.16 at 5:00 pm

Hi Harry. Good for you! Please don’t be too hard on yourself. I very much doubt you’d permit any of your students to judge their efforts to be ‘failures’ the first time out. I mentioned before that I record my own discussion classes to assess pedagogies and productive output. Believe it, or not, you’ve already made immense progress by making that first step of bringing the cameras into class. Students are operating in different space and both they and you are going to change your behaviors for the better (almost certainly).

You’ve done the most important work already in assessing the kinds of exchanges taking place. I’m a big fan of starting with very simple metrics. What kinds of output are students producing? Who initiates discussion – do the students use questions, or just answers – head knodding, or actuall utterances.

Then we move onto content. How often are students referencing other student comments and the assigned readings? Are students recording other student contributions in note form? Are students using paper and pencil to write down new questions for other students during the discussion.

Now mechanics – critical as you suggest. An hour, or ninety minutes is a very short time for a well-organized discussion class, or a very long time for unstructured discussion. Many good teachers I know break discussion classes into small groups of 4, or 5. My own preferred size is just 2. Monitoring 8-12 pairs of 2 is much simpler and effective. Question formation needs to be an assigned part of the class and is, as you know, a metric that provides real insights into student grasp of issues. Students are required to restate the ideas of others and comment on these positions. Students understand that much of their assessment is not based on what they bring to class, but rather how well they respond intelligently and creatively to what others bring to the class.

Affective barriers play a very large role in these classes, especially for the less confident students. Laughter is essential. I encourage my students to laugh at me, not with me. Pointing to my own foibles and shortcomings in the discussion class is a great way to begin the change/development process. Start with: ‘Our first class was surprisingly good in many way (Everyone talked and contributed. Hooray! That is measurable, positive change.) And, we can all do better. Build from success with clear goals over ten weeks for example. Graded assessments each of 10 weeks (70 percent) and final assessments (30 percent).

Time constraints. I normally have students work together in pairs under time constraints. Five ten minute discussions with five different partners. 3-5 minutes of presentation lead each, followed by 5 minutes discussion. Students prepare very short presentations before class. These can be on assigned questions/or student-selected material connected to assigned themes. (the former is usually better, but both are necessary)

Students summarize and report what they’re learning from each student to each other and to me. It works like a charm, frankly. My own contributions to the discussion are uniformly positive – ‘Great question! Interesting connection, comparison. Great answer!, now please ask your partner a good simple question.’ The multiple partners break the class into manageable units. With limited time constraints, the windy are forced to become concise and focused. A variety of partners in the same class allows all students to both respond to multiple different ideas, and to hone and refine their own messages. Students normally get 3 minutes to modify their presentations (which usually involve forming questions for interlocutors.

Many teachers I know use this system. My own modification to the system is to keep the discussions to pairs. That way I can identify all the dominant/passive students immediately and immediately deploy strategies to redress inequalities in production.

Organization usually involves me directing the transitions very quickly to prevent students gravitating towards/away from specific students. Usually an A/B split for the pairs. All As remain in place, all Bs migrate one place clockwise. Discussions begin semi-formally with salutations, names, but move very quickly into lead/audience work.

As noted, students love the work for the most part. Students produce work that can be monitored over the course of 70-80 minutes and over 10-13 weeks. It’s pain-free teaching and produces results that are provably superior to other approaches, imho, simply because we have clear, concrete records every week of every student’s performance in activities that are replicated over ten weeks. There are numerous opportunities for modification as each teacher determines.

Finally, one key point. A mobile devices and electronics are turned off and placed in student bags at the back of the class. All extraneous materials are banned from the discussion space. Students bring sheet paper, notes, pencils (a beverage) and their energy to the discussion space. After the first class, I record discussions at several points during the term, not every class.

To conclude: kudos, Harry!


kidneystones 09.12.16 at 5:01 pm

knodding? So me.


kidneystones 09.12.16 at 5:06 pm

And one note for content clarity: ‘ Students normally get 3 minutes to modify their presentations (which usually involve forming questions for interlocutors.’ after/before each partner change. I usually monitor this activity very closely.

And, beginnings and endings make each task more manageable for the less-confident. Students can and are expected to improve the quality of the delivery of their arguments over the change of five partners.

Looking forward to reading other comments.


Philip 09.12.16 at 7:09 pm

Reading the OP, I was thinking along the lines of passer-by in that you never explicitly stated the advantages of more student led discussion over tutor (expert) led discussion. My teacher training was as an EFL teacher so it was drummed in to keep teacher talk time to give students more opportunity to practise using English. I guess a similar thing could be working here i.e. you want students to practise discussion skills, making an argument, responding, questioning etc. and also that this will help internalise the content and therefore aid learning.

Another point from EFL is accuracy vs. fluency so a common practice is to make notes of errors made in discussion and then present them at the end (possibly modified so as not to single out people for making mistakes) and make an exercise to correct the errors. So a better question than do I need to interrupt might be do I need to interrupt now. So your final summary might be the best point to elaborate on areas that weren’t discussed much or to clarify possible confusion. Also as kidneystones says giving students chance to discuss things in small groups before whole class discussion or feedback gives people chance to articulate ideas first so hopefully they will be less reticent to speak or you can monitor the smaller groups and maybe hear what someone has to say who won’t speak out in front of the whole class. However for that to work you need to make sure the question will engage the students either by knowing it is on a topic they care about or it is clearly linked to learning about the content. It is also necessary to think why do you want this discussion, at times it can feel like the tutor is just using up time. So is it to practise skills, to make students engage with the content, or to check their understanding to inform what to cover in future sessions and where to pitch it?


PatinIowa 09.12.16 at 7:55 pm

I highly recommend this:

Tompkins, Jane, “Pedagogy of the Distressed,” College English, Vol. 52, No. 6 (October
1990), pp. 653–660.

You–probably many of us–are doing many of the things she advises.

I have my students read it, too, and explain that I’m trying to get them to talk to each other as much as to me. (I teach small group composition and gender studies classes; I’m lucky to have formats and classrooms that are amenable to this approach.)

I just came from a class where I was (sort of) able to discipline myself (it’s difficult; I’m still an honors student), and while it wasn’t perfect, they did carry the load admirably.


Eszter 09.12.16 at 8:46 pm

I love discussing teaching with people who care about teaching and have lots of thoughts on the topic of discussions so please forgive the marathon comment.

This past Spring, I taught a seminar that students seemed to love (judging from their in-class behavior and also their anonymous evaluations). It had 12 students and it was a completely new course, not just for me, but seemingly for anyone out there. It was an undergraduate seminar on the social aspects of algorithms and I had to come up with the syllabus from scratch. As you can probably tell, it’s a fresh topic, an evolving topic, so I tend to have as many questions as answers. I believe that it is important to make that clear to students and show them that we are doing learning along the way with them, if that is indeed the case.

This was the first class where I had undergraduate students lead class discussions. It was an upper-level seminar and most had taken other courses with me so I knew how high their caliber was and I knew they could handle it. I had students lead six classes, they knew the topics, I asked them for their preferences, and managed to get everyone one of their top two choices. They were in pairs for discussion leading. They had to send me their slides about 24 hours before class for feedback.

I went back to read what were stellar evaluations (for once without any negativity) to see what worked for students. Here’s one comment: “A three hour class flew by because of the quality of the discussions, both led by students and guided by her.” I thought this phrasing of “guided by” was interesting. That is indeed what I was often doing. I was guiding before class through my feedback on the slides. And I was also guiding during class. Sometimes the discussion leads would make comments to which there was no reaction and they were about to move on. But then I would ask them to hold off for a bit and explicitly asked the class to react. It is not surprising that undergraduate students have a hard time getting their peers to contribute to class discussion given that it’s a challenge for many seasoned instructors. So on that point I was constantly there ready to jump in with a nudge. Sometimes my questions were fairly open-ended to ask for more general reactions. But sometimes they were specific about points the leads were making and asked whether people agreed with the arguments presenting.

Generally speaking, asking whether people agree with a comment that was just made can be a good way to get students to talk. Be sure not to reserve that question only for times when you disagree with the comment, that could become predictable.

As noted above by M Caswell, it’s helpful if students know each others’ names. What I’ve found surprising is that even when we have table tents out with people’s names (all of my classes at NU did) and I refer to students by their names, they don’t do so with each other. I’ve never explicitly mentioned this, I wish I had brought this up at some point. This past Spring, I did ask in Week 2 or 3 if anyone could name everybody else and one person almost got all names. She managed by the following week. Perhaps this emphasis on knowing people’s name helped students learn each other’s names.

In my experience, students start referring to each other by name and having more actual discussions referencing other people’s (other than mine, that is) comments when they have already taken another class with the other students (and possibly me as well so that the overall dynamic is familiar). This is, of course, relatively rare, although I did end up having several students who had taken several courses with me so they definitely knew each other. In such a situation, of course you have to make sure that it doesn’t end up seeming like a clique that is not open to newcomers. Fortunately, I don’t think that ever happened.

What has worked very well for me to make sure students are ready for discussion is requiring a written memo as homework assignment sent to me the evening before every class. This is not particularly scalable, however, and can be quite a bit of weekly work for the instructor even with 15 students (my cap for seminars at NU), not to mention larger classes. But it can give you an excellent sense of what students are getting, what they are missing, and what interesting points the readings have raised for them that you didn’t even think of. With everyone having done the readings enough to write a memo on them, people are usually ready for discussion.

Also, I have no problem with silences, which is an absolute must for encouraging discussion – especially without resorting to calling on specific people. I hate cold calling on people as that was the kind of thing that petrified me as a student. However, as noted by someone else earlier, I do look around the room carefully to see whose facial expression looks like they have something to say, but are just too shy to raise their hand. I then look at the student longer than usual and offer an encouraging look, which often gets them to open up. If not then I may actually call on them, but move on along if I misread the facial expression (pretty rare, actually). I think it helps to establish a dynamic where it’s okay to say that you don’t have anything to add at the moment since it could help in reducing comments for the sake of comments.

I have a question for passer-by since I’m about to start teaching in Europe and am discouraged by the lack of discussion here. Would it have helped if the professor had explained why it is beneficial to have discussions even if you have to listen to some students who may not know that much more than you do (or possibly even less)? One thing I do not do in my classes much is let students talk among themselves. I do it occasionally briefly, but if I can’t be there to hear what people are saying and eventually react then I consider it not a great use of time. Interestingly, students in the US always ask for more of it, but I’m not sure that type of discussion is necessarily that beneficial. There are exceptions, to be sure, but it’s not something I’ve pursued much. That said, I may need to as I start having larger classes and still hope to have discussion time.


Metatone 09.12.16 at 9:41 pm

@Phil – are there fixed tables in the room too?


Val 09.12.16 at 10:21 pm

I too love these teaching posts Harry! I think that we as teachers learn from them in the same kind of way your students do, to some extent.

I’m trying to finish my thesis so am only teaching in the one postgrad online unit now – also interesting but different issues, as it’s easier to get mature students to have good discussions anyway – but if and when I go back to undergraduate teaching I really look forward to trying some of these ideas out. We use s lot of group work so teacher led discussions aren’t such an issue, but there are times when ideas are talked about in the full group and I really like yours and some of the other ideas here about when and how you might “guide” discussion or “lead” it – or back right off at times.


harry b 09.12.16 at 10:27 pm

Thanks for the kind words, I really appreciate them, and it does prompt me to put more energy and effort into these posts, knowing that they get read and some people find them helpful.

Talking of which I have been drafting a post on why discussion is important, and passerby’s and Phillip’s responses have spurred me to return to it! I’ll try to get it in within 2 weeks…

Thanks both for the kinds words, and the details, kidneystones, that’s all very useful and helpful.

But…. Ok, I am not going to rant, but ‘fixed chairs’!! We have a horrible building (old, mold, leaks from the ceiling, floors that never get cleaned, etc) which I opt to teach in because it still has smallish classrooms without fixed tables or chairs!! It’s paradise. Two rooms have been refurbished, at a cost which one architect in the know estimated at $50k, with fixed chairs! They have spent $100k making those rooms impossible to learn in!

My advice is to ask for a new room, and be specific about what you want (movable chairs), and that you will put up with a lot of crap to get it. Choosing unpopular teaching times helps with space — on my campus, we have much better choice of rooms at 4pm on Tuesdays than at 1 pm on Tuesdays.

Loads more to say, so I’ll try and find time later…


Alan White 09.12.16 at 10:31 pm

Harry, as always thanks for pushing matters of undergrad pedagogy to the fore in thoughtful and practical ways.

As to one very practical matter here: how do you get all or nearly all your students to read assignments, presumably necessary to participating in discussions? Is a warning about participation enough? My experience is that reading quizzes with a trifle of credit help, but typically even there those reveal only about a 60% preparation rate.


harry b 09.12.16 at 10:40 pm


kidneystones 09.13.16 at 1:09 am

@15 Hi Alan,

Peer pressure plays a key part. Separating the students into two groups: those prepared and those not ensures that those who bring nothing to offer to the class are not inflicting their shortcomings on others. Compelling students to restate the arguments and questions of others ensures, frankly, that very few students walk into the classroom without being able to contribute. That’s one of the reasons I compel students to work in pairs with a short preparation sheet and with references to assigned material in the discussion. The second is the grading system clearly identified in the syllabus and reviewed in the first class. Class behavior is just like a part-time job. No work – no pay. That’s a concept all understand. Those unwilling to adjust to the class drop out, but given the fairly simple structure of the class, attrition rates are very low. Asking students to read 30 pages of anything for a discussion class is asking a lot. Asking first-year students to have read 5-15 pages, 5 being ideal, puts students on the success track. Reading loads increase over the course of the term.

The key is to build success into every activity in every class. Strong students shine out of the gate, but given that 50 percent of the discussion grade is based on how well they involve their partners respond to their arguments they, too, must find a register that works, even for the less-involved and less-engaged. They can’t simply stand back and show off. They get they’re grade on the quality of the re-statement of their ideas produced by their interlocutor. eg. Me to B: how much of the A’s argument did you understand? Percentage? What does A think about ? Why? Was their a counter-argument? Me to A, ‘is this right? What questions did B ask you?

The students get it, really. And as long as the readings are short for the first week, most will respond. I normally allow students to choose from a very large selection.


kidneystones 09.13.16 at 1:11 am

Lots of typical typos, sorry.


Alan White 09.13.16 at 1:30 am

Thanks Harry. You’re right–getting students familiar with one another is a big part of encouraging participation. But in the previous post you mention at the end that you would try these same kinds of strategy for larger sections. Were you able to do that?

kidneystones–you’re right about getting students invested in the material, and especially in the beginning making that investment manageable. I never assign onerous readings (well, almost never–I still assign Fear and Trembling in Phi/Rel), and I emphasize comprehension over content. I will confess that most of my career I have been more a lecturer than discussion leader, but always keenly aware that student contributions are essential to useful learning. The one class out of thousands that haunts me is one where I very cruelly shut down one student for seeming to dispute my judgment. It’s the one day in my 35+ years I wish I could do over.


Peter Dorman 09.13.16 at 3:38 am

I teach at Evergreen State College, where seminars based on readings (usually books) are considered essential to the pedagogy. The founders of the college in the early 70s had two main inspirations, Dewey and Meiklejohn, and the seminar comes to us via Meiklejohn. (Although Dewey had interesting things to say about discussion in an educational setting.) So no class (we call them “programs”) is complete without lots of seminars. Faculty think they are the crowning glory of the college. Talk to any student, and you will hear mostly complaints about wasted time—but not always. We typically have 20-25 students per seminar group, which may be part of a larger learning community that gets together for workshops, lectures and other activities. Full-time programs meet 12-16 hours per week, of which maybe 3-4 will be seminar.

I’ve been teaching at Evergreen for 18 years. (I taught at many other schools previously.) Here are three things I do that might be useful for faculty newbies. (1) In the first seminar I verbally ask students to talk to the whole group and not just me when I see that’s what they’re doing. I accompany this with an “all of us” hand gesture. In subsequent seminars I can use just the gesture. (2) I have student facilitation for most seminars, but I make sure to meet with each facilitation group in my office at least a day before they’re on in order to discuss their approach with them. One reason I do this is that, as they’re preparing for their big event, they’ll be watching the seminar closely to see what’s working and what isn’t, and of course I want all students to see the process through that lens. It also makes it much easier for me to participate in discussion as a relatively equal member, relieved of the responsibility for managing it. (3) When I’m facilitating, which I do at the beginning of each quarter, at strategic moments I call a time out and ask students to write down their thoughts. When we resume I will do a bit of cold calling to get things going, but it is less terrifying now since there’s something on paper to present. The purposes are to up the level of reflection and especially to get more participation from quieter or more deliberate students whose points of view would otherwise be underrepresented.

Nevertheless, I have two big issues I haven’t resolved yet. The first is that the subject matter I tend to teach (political economy, environmental science, history, etc.) has a right-wrong boundary. If a student makes an argument that rests on a misunderstanding of the balance of payments accounts, for instance, that needs an intervention of some sort. My intervening on the spot ruins the dynamic, but waiting until later can be insufficient correction, since one argument builds on another, and the next thing you know you have a parallel universe with laws and principles quite different from the one we inhabit. The second is that seminars often fail to get to the key questions about a reading. These are typically lodged in the assumptions underlying it, and students tend to adopt the author’s assumptions without being aware of it. To some extent I can get around this by being pretty directive with facilitators in our pre-seminar meetings, but my experience is that this falls well short of what I’d like. I’ll admit I find it difficult to control myself during discussion if I feel like we’re circling around the core issues and not engaging them.

One passing phrase from the OP that struck a nerve with me is “though where’s the rigorous evidence?” Yes. We need more assessment, and “we” includes the student participants. One thought that occurs to me is to have each student, at the end of each seminar, write down the most important things they’ve learned during the past hour or two. I might go so far as to direct them by saying that what they write should fall into one of three categories: a wider view (seeing new perspectives, arguments or concepts), clarifying the obscure (moving from not understanding something important to understanding it), or complexifying the assumed (moving from a superficial clarity to a complex web of deeper questions). My guess is that any really significant takeaway belongs in one or more of those three boxes. Then, at the end of the quarter, I will ask them to review what they’ve noted along the way, revise and consolidate as necessary, and turn in a single list. This would help me write their evaluations too. (We have narrative evaluations instead of grades.) I will confer with my teaching partner about this.


magari 09.13.16 at 4:54 am

Kidneystones, I’m not sure I totally get what you’re trying to say, but I would like to since I’m teaching a seminar next semester and am always looking for new techniques to better student involvement.

(Part 1) So you break the class into groups of two and each group is responsible for doing something (Person A gives a “presentation” to Person B and then they discuss?) and then they rotate (Person B gives a presentation to Person A? or Person A gives a presentation to Person C? And what about Person B? Is she also giving a presentation to someone or some group of people? Is every person giving a presentation to some other student or group of students in every class meeting?).

(Part 2) Students report what they’ve learned to you. What does this report consist of, and do they do it after each and every presentation? Or after the entire “round” of presentations?

(Part 3) You give praise. And constructive criticism, I presume? Teachable moments, and whatnot? When do you activate (or how do you build in) your role as an instructor?

(Part 4) What are you doing during these student-student interactions? Say you have 10 groups (20 students/groups of 2). How do you evaluate student performance given there are 10 groups operating concomitantly? And how do you evaluate them given that your presence may act as a disturbance?

Also, to Harry, these posts are great. I’m looking forward to your future note on the reasons for the importance of student-student interaction. My primary worry in lionizing student discussion and student-lead discussion is that it primarily acts to increase their rhetorical capacities–their ability to think on the fly and convince other human beings–without necessarily increasing their knowledge of the material and ability to think behind and through it.


kidneystones 09.13.16 at 5:48 am

Hi Magari, Thanks for this and thanks to Harry, again, for the forum. Apologies for the lack of clarity.

You’ve basically got it.

1a/ All students select a ‘presentation’ topic from selected readings. In some cases students select exactly the same topic. Students are expected to study their reading closely, and read any extra materials necessary to ensure they can put the reading in context for the audience.

1b/ Pairs are divided into two groups: One student in each pair is an A and the other a B. For visual purposes imagine a clock. All students designated A take a place at each hour of the clock and remain at that point throughout the class. All students designated B will move to a new partner around the clock one place. All B students move in the same direction at the same time, with appropriate thank yous, that was interesting etc.

1c/Each student in each pair presents for 3 minutes on the reading/topic. The student listening is expected to be able to formulate several simple questions to elicit nuance, clarity, and detail. The engage in a short discussion and then change roles.

1d/ At the transition point students review their work, make changes for clarity, and make notes about the preceding discussion.

2a/ Summarizing and restating. This is done orally. Usually, at the transition point, or during the discussion. You’re right. I am intruding, but principally to confirm that both students are actually on task. I’m not particularly interested in the content, but more in the form of discourse. Are both students contributing ideas?

2b/ Evaluation. I collect all the presentation notes and documentation at the end of each class to see what students brought to the class, and how they modified their presentation notes over the course of the discussions. Pristine pages are far less preferable to a page covered in arrows, hastily scribbled additions, underlines, scratch outs, exclamation points. I photograph these as permanent records of their class work. I hold regular consultations with each student for detailed two-way feedback. I’m principally interested in construction of arguments, erecting counter-arguments and responding to the changing dynamic. The point? A discussion should change the way I think about my own topic, reading selection. Students are asked to explain how their understanding of the topic changed during the class activities. All have clearly become invested in the project and expect to have a more nuanced understanding of the material.

3/ I think this had largely been dealt with in points 1 and 2. I don’t see myself as teaching a truth, but rather helping students become aware of and participate in the larger conversation of their topic. I encourage them to search out other sources to increase their knowledge. Those who do, obviously, receive better grades.

4/ Monitoring is simplified by ensuring students stand in pairs about a meter apart from each other. I constantly move around the room listening and observing. As the students are repeating and honing the same presentation five times, I usually get an opportunity to hear each presentation in its original and modified/improved form. I say very little at this stage. I intercede only/usually when needed, or when I hear an especially salient point that I wish to bring to the attention of the class. All student work is kept in a A4 folder. I collect the folders at regular intervals to monitor what students are actually producing on their with homework assignments and other work.

5. Strong students identify themselves very early. Student with the capability to become strong can be encouraged to compete for top grades. The baseline grade is B for the first 4 weeks to ensure all students understand the tasks and have the requisite skills to produce material for evaluation. I only award a handful of A’s. When teaching research skills, using discussion methods, students usually work the entire term towards a single presentation. All the same activities of discussion, retelling, forming questions are part and parcel of these classes, too.

Finally, after a long comment, we need to recall that the discussion is only one component of the class, perhaps the most important. I don’t ‘teach’ students anything. I introduce them to discussions and provide them with the basic skills to ensure they understand the key contours of the larger conversation and have the equipment to evaluate arguments for themselves. The video makes it all real and provides me with something far more reliable than my mere impressions week to week. Regular one-on-one meetings with each student to ensure they grasp key points occur 2-3 times a term.

Good luck, hope this helps.


passer-by 09.13.16 at 8:30 am

To Eszter: although teaching cultures vary a bit accross Europe and accross different institutions, as far as I can tell, there is very little real class discussion in continental European university classrooms (do not know about the UK though). Here, the challenge for those who want to encourage more active student engagement in classes is much greater, and you have to go at it gradually.
Yes, you absolutely, definitely have to make a very cogent defense of the value of discussion for learning. You have to keep in mind that overwhelmingly, students will have very little positive experience with such experiments; most of their teachers do not have the skills or the interest to encourage class discussion. I have studied and taught in three different countries, and systematically, when “participation” was said to count for the grade, the students ignored it as a joke. It’s usually the kind of useless fuzzy rubric that the students suspect allow the teacher to freely adjust grades as he feels, or it means that the students only vye for the teacher’s attention. Do NOT try to tell them that their participation in the class of 30-50 students meeting once a week will really count; they are not stupid: either it does not really count (because you cannot meaningfully evaluate that) or it means that you are giving out bonuses to the few students who will have managed to catch your attention in that setting.

Most of the students are also not used to actually speaking up; the ones who do are not only the more self-confident, they are also very often the ones who are both academically strong and trained in much more traditional school systems (emphasis both on competition and rhetorics – even if nobody calls it that). When there are discussions, those mostly take place in elite institutions / groups which tend to favour extremely competitive cultures. For both those reasons, discussions, when they do take place (not the lazy answers to satisfy the teacher and get them to move on), tend to quickly veer more toward competition and rhetorical sparing then toward actual collaborative learning (which, I assume, is the goal).
To stop that dynamic and ensure everyone’s participation, mere techniques (as the ones offered in the comments – thanks a lot!!!!) do not suffice. You absolutely have to convince all students, especially the “best” ones / those in more elite settings, of the academic value of such discussions. If they don’t buy into it and consider the whole exercise a waste of time / a fun easy class, it is impossible to achieve any kind of learning through discussion.


passer-by 09.13.16 at 8:49 am

Oh and one last point that I haven’t seen addressed at all, is the impact of grading systems on class discussions.
I had the impression (maybe biased because of the Ivy League setting) that in the US, grading was much less punitive / much more generous than in Europe. As a rule, universities in continental Europe do not pre-select their students (or only a very weak selection) and they are free (so not student customers).
Not only is there no pressure toward grade inflation, on the contrary, many universities actively use grades to select the students during their studies – culling them. At the university where I taught as a graduate TA (in history), the goal set by the university and professors (even though not publicly acknowledged) was to pass half the students for each of the three bachelor years (yes, we were actively curving the grades to get a 50% fail rate). Both in France and Germany (the two countries I am most familiar with), the prestige / reputed quality of a course or institution is routinely discussed in terms of how many students fail at it (proving that it is “hard”, thus “good”).
It is very difficult to convince students of the value of class discussions when they are under such pressure (no wasted time!) and to foster the kind of trust necessary for the discussion to be productive when the students know that half of them will fail the class / no more than 10% will receive an “A” or equivalent. And no single teacher can fight that kind of culture alone.


TM 09.13.16 at 10:04 am

25: “pass half the students for each of the three bachelor years”. This seems quite an exaggeration, at leats wrt Germany. Usually they put the bar high in the first year and many students do fail the first exams, but even in Mathematics, where the fail rate is definitely high (which doesn’t in my experience convey prestige the way selective admittance rates in the US convey prestige), I’ve never heard of a fail rate of 7/8 over three years.

Eszter: Curious what your experience will be. My guess is you are in for a culture shock.


passer-by 09.13.16 at 10:11 am

@TM: that was a main Paris university. It is a bit complicated, because the French system is absurd, but that particular (“prestigious”) university “clears” the ranks of the students and makes up for it with an influx of “better” students coming in in the 3rd year and beyond from the competitive “classes préparatoires”. Less than


passer-by 09.13.16 at 10:37 am

Sorry for that.
@TM: that was a main Paris university. It is a bit complicated, because the French system is absurd, but that particular (“prestigious”) university “clears” the ranks of the students and makes up for it with an influx of “better” students coming in in the 3rd year and beyond from the competitive “classes préparatoires”. At that institution, we used to graduate less than a third of the students matriculating out of high school (and the professors were really unapologetic about it – I still consider it an shocking and abysmal failure). It is extreme, but high fail rates are really not that unusual in European universities.
In Germany, a third of university students fail overall at universities; the quote goes up to 50% in engineering at universities. As far as I can tell, it’s the same in math. Some institutions choose to focus the culling on the first year (that’s the case in medicine in France, and in Germany as well, I guess – it’s too expensive for the state to invest that much in students who drop out); others do it more gradually.
But statistics aside, I do get the impression that there is a strong cultural / systemic difference that makes it very difficult to foster the kind of collaborative learning environment discussed here. Even if the selection pressure lessens the farther a student goes in his studies, it is not easy to convince them that the strategies that ensured their success up to then and allowed them to survive the selection (competition, focus on grades, traditional learning skills in lecture-heavy environments, skills at exam writing…) no longer suffice / can really be complemented by other skills and qualities, other teaching environments.


Josh mane 09.13.16 at 11:34 am

As noted, students love the work for the most part. Students produce work that can be monitored over the course of 70-80 minutes and over 10-13 weeks. It’s pain-free teaching and produces results that are provably superior to other approaches, imho, simply because we have clear, concrete records every week of every student’s performance in activities that are replicated over ten weeks. There are numerous opportunities for modification as each teacher determines.


James Wimberley 09.13.16 at 12:43 pm

Phil #3: fixed seats are a disaster, not just for teaching in any form other than the formal lecture, but for work meetings generally. Teachers have to insist on flexible furniture: seating is part of their pedagogical responsibility.

Francis Bacon is typically astute on the importance of seating round a table (essay Of Counsel.) My experience has been of international meetings, not teaching. We were constantly hampered by meeting rooms designed by architects with little understanding of the work, briefed by similarly uninformed technical services people.

Live interpretation is vital to international work, but it creates its own problems. The seats need microphones and headphones. Only recently has wifi technology got good enough to allow complete flexibility of tables and seating. In a meeting of any size, the interpreters are in cabins. They need sight lines to the participants, to see their gestures as well as hear their words. That is compromised by circular seating. The use of microphones limits spontaneity; the chair has to control the order of speakers. Interruptions are basically disallowed: a blessing in disguise.


roberto ferreira 09.13.16 at 1:18 pm



TM 09.13.16 at 1:45 pm

28: “In Germany, a third of university students fail overall at universities”

To my knowledge, there aren’t very reliable data on drop-out and graduation rates in Germany. One third could be right but it’s probably a guess and may or may not account for students who change their degree program. At US universities, comprehensive statistics on graduation rates are available for each institution and the average *six-year* graduation rate at colleges is 60% (58% at public institutions,; again this probably doesn’t account for transfers). I agree that the role of grading is different – US students fail their required exams much less frequently but many still drop out for other reasons.

In my experience, in Germany the “selection pressure” so to speak is clearly concentrated in the first year or two of studies and the atmosphere is rather different after that. But that was before the Bologna reforms which shortened the time allotted to students, imposed much more structure on study programs (following the anglo-American model btw) and thereby increased pressure quite a bit.


magari 09.13.16 at 1:58 pm

Thanks Kidneystones for the clarifications. So if your students are doing all this standing up, I guess that works around the fixed seat problem!


kidneystones 09.13.16 at 3:33 pm

@ 32 Cheers, magari. Yes, and it’s much easier for everyone to stay awake.


TM 09.13.16 at 3:34 pm

28: Re curving the grades, I can’t speak to France but in Germany, I have never experienced that either in school or Uni. I would expect that any self-respecting teacher insists that the bar is set in advance and every student exactly gets the grade they earned.

Re selection pressure, at least in my time there wasn’t a feeling of being in competition with fellow students. We knew the standards were high but it wasn’t that you had to be better than the other students (or better than 50% of whatever); you had to meet the standards.


Tracy Lightcap 09.13.16 at 3:54 pm

Some further suggestions.

First, I sometimes arrange discussions. The method is to divide the class into groups who have to be ready for a discussion of a particular body of material. Then I use a random number table to pick one of the groups. They then move into a center circle with the rest of the class arranged in a wider circle around them by group. The chosen group gets 20 – 25 minutes to discuss the assignment. Then the discussion widens to the outer circle where a critique of the discussion by the center group is expected. (I’m careful to point out to all my classes that a critique can be either negative or positive; praise for good points is as acceptable as criticism of shortcomings.) This method keeps everybody on their toes alright and the groups enforce readiness on their own. It can also lead to students dropping the class; many are uncomfortable with this level of commitment. Thus, I tend to keep such discussions to the exact number of groups in the class (6 groups, 6 discussions) and give everybody plenty of advance warning.

Second, I use a lot of “guided discussions”. In this the students have an assignment to read and think about. When they get to class, they are given a worksheet with instructions to write out a question and a one sentence explanation on their own. (For example, I use The Moon Is a Harsh Mistress as a reading in one of my classes. The question: Write out the most serious question that a new chum would have for a seasoned Loonie. The explanation: An earthworm has just challenged you to give a one sentence capsule of the principles guiding the new Free Luna government. What’s your answer?) I then put them in groups (I change the composition of these each time.) and require them to discuss their questions and explanations and come up with a question and explanation the group can agree on. (They don’t have to use the questions and explanations they come to the group with; they get editorial discretion.) They then go to the board and write down their question and explanation. I then ask the question of another group (“You Loonies in group three: what’s your answer to the New Chum?”) – or ask them if the explanation holds water. The group that puts the material on the board gets to chime in if there’s a disagreement. If the questions and explanations are similar I will sometimes ask a particular group to justify their version. (I’m the Group Overall Director – G.O.D. – after all.) This can take a full 90 minute class and gets everyone engaged. The next class is usually a lecture where I try to put the discussion into overall context.

I also do just plain jive class discussions, but after using one of these techniques. Once the students get used to contradicting me and thinking on their own, the usual model works better.

One other trick I have found useful is the two minute drill at the start of each class. I get the students to write out any questions that arose last time about the material covered and try to answer the queries. A good device here is to get them to write out the questions on 3X5 cards and treat the responses anonymously. I try to find one that either a) has a thought I hadn’t come up with (“That’s interesting. I’ll steal the idea, if it’s ok with you.”) or a point I forgot to cover (“Doh. My bad. I should have said this …”). Discussions are easy to start from this basis.


Eli Rabett 09.13.16 at 11:18 pm

Tracy, that has important elements of POGIL. the groups should be small, three or four. Each student should have a role, manager, recorder, reporter and after discussion they report their conclusions and there is a general discussion.


Michael O'Hare 09.14.16 at 3:41 am

Interesting discussion. More and more of us are realizing that acquiring content is best done from the web or a textbook, so what used to be lecture time is better spent actually using that content. James’ point about classroom design is very important. Luckily, classrooms are a very small fraction of the cost of teaching, so there’s no excuse not to have suitable ones.;2-O/abstract
On grading, in classes up to 90 I have had excellent results in attendance, quality of engagement, and everything else by letting the students grade each other on class participation.
When I’ve taught in France and Italy, I had the sense that students were pathetically grateful to be treated like grownups and engaged in serious discussion.


passer-by 09.14.16 at 8:50 am

Michael, thank you a lot for the peer grading suggestion and very informative link. That sounds like an excellent strategy indeed.
French undergrads are, indeed, treated as schoolchildren in teaching (and if you’ve been to a French school, you know that that is NOT a good idea) while also expected to be completely independent and sink or swim in an university environment that provides little to no support. The best results I have had came, indeed, from treating them as adults. It just so often runs cross-purpose to what the institution expects from them.

FWIW, German undergrads are probably among the most mature and independent students I have worked with. Problems come more from the environment, professors and institutions – too many exams, too little time, too big classes, no support…

TM, the Hochschul Informations System and the DZHW ( do some very good work on the difficult problem of drop-outs in Germany. They are the ones who estimate that a third of the students leave without a diploma (distinct from the 10% or so who switch programs or universities), up to 40-50% in math or engineering.


TM 09.14.16 at 10:10 am

Thanks for the link. I’m always interested in these kinds of statistics. The paper gives the overall dropout rate for bachelor porgams as 28%. This was based on students beginning in 2008 and if I understood right, the study goes up to 2012. So 72% would be the four year graduation rate. But I could be misinterpreting, the methodological explanation could be clearer.

What I would like to see is an in depth study that looks at cohorts say in their 30s and 40s and documents in detail their educational history: how many have a university degree or some other kind of educational certificate? How many took up tertiary studies, and how many of them completed the degree, how many switched programs with or without success, those who dropped out: why, and after how long, and what did they do after dropping out, etc. etc. etc. These kinds of data are hard to come by, I don’t know why. But they would be extremely interesting.


passer-by 09.14.16 at 11:46 am

You’ll find some info here, TM: (link to full report is there; with explanations as to why the data is so bad).
Germany does do a great job at leading their high school graduates to qualification and work. Universities, however, are probably the weakest at that – the Fachhochschulen (“universities of applied science”, I think, is the translation – high-level academic education with a focus on Praxis) do a much, much better job. That’s why you come to an overall 28% drop-out rate: the FH’s are way better at retention than the universities. I do hope that they won’t mess up the excellent FHs with the Bologna reforms…
Anyway, in Germany at least, dropping out of university is in no way a disaster for the students, which may explain why the episodic discussions of student retention lack urgency. What bothers me is that quite a few professors use that as an excuse for not really trying to improve their teaching and outcomes. As a teacher, I do want to do the best possible job at getting my students to learn, even if in the end, it’s no big deal for most of them if I fail at it.

PS: my personal experience with the German system was mostly during my PhD / in post-doc positions, during which I only taught two classes (but was invited to lead quite a few seminars), and, for various reasons, I still work very closely with German institutions. My sister-in-law, however, is a German economist doing research on education. So a lot of what I “know” comes from discussions with her, so I may misremember quite a bit.


Meredith 09.15.16 at 4:37 am

A silly little practicum in the midst of the fine discussion here. To get students talking to one another. One student responds to teacher’s initiating-discussion question. A second student responds, probably looking at teacher rather than initiating student. Teacher: look at second respondent, of course, but keep your eyes mainly on first responder. This body language communicates to second responder: you’re responding to her/him, not me. Teacher’s eye direction (and other body language) says a lot to students.


harry b 09.15.16 at 1:26 pm

“A silly little practicum…”!!
That’s exactly the sort of specific advice that we all need! I think I do pretty much exactly what you say, and it does play an important role, but its interesting that it is not in my head enough to have explicitly said it in the OP (and I will, now, if I present this in other fora). Of course, this is a perfect illustration of the value of discussion!


Tracy Lightcap 09.15.16 at 9:01 pm


It is a bit. The template I used for this pre-dates the POGIL project by some years.

The main differences come in the character of the group work. There are no group roles in the guided discussions I use and the amount of written work is both less and different. That’s to be expected in a social science (I’m a political scientist) class, at least for this kind of exercise.

I do something similar to POGIL occasionally as well. I have the students get in groups and access STUDENT CHIP ( on the internet. They then use the datasets there to work out a series of simple descriptive and bivariate analyses. Then they compare their results with other groups. But this doesn’t happen that often; I usually assign the exercises individually.

Comments on this entry are closed.