Generating More Effective Classroom Discussions

January 20, 2003
Presented by Ken Jones

Let's go to an ideal world for a moment. What impact do you hope to have on those students in your classes?   Do you want them to retain the material after the course is over?   Do you want them to be able to transfer that knowledge to new situations?  Do you want them to have improved problem solving skills?  Do you want them to be inspired to learn more?

These are all pretty marvelous goals and ones I'd certainly be happy with.  As it turns out, educational research over the last 25-30 years has repeatedly shown that discussion based courses provide better results than lecture based courses in all of these categories.  So why then do we talk so much?

I'm sure there are lots of reasons why so many of us continue to lecture, but let me suggest four.  The first is simply inertia, or to be a little nicer, we continue the practices that we saw modeled when we were students.  Second, if you are like me, you simply love to talk about your material, and you assume that they like listening.  Third, we tend to segregate roles; ours is to provide the information while theirs is to learn it.  Lecturing allows us to uphold our part by "covering" the material.  Fourth, discussion is simply harder.  With lectures, all we need to do is tell people something we already know.  With discussion, we need to figure out how to help them learn that knowledge.  So we have "telling" versus "facilitating learning."   They are very different, and the latter is much harder.

Good discussions are spontaneous and unpredictable, but there are lots of things we can do to in advance to increase the likelihood that good things will happen.  Let's start with what we can do as faculty.

Faculty Planning

1.    Choose meaningful and appropriate things to discuss.  You can't have a good discussion on questions that have a single obvious answer or which are completely outside the knowledge base of the students.  Good topics can be drawn from controversial issues, conflicting viewpoints, case studies, or key concepts and theories (rather than just explaining how a chemical reaction occurs, you can discuss it if you ask why it has to work that particular way and not some other.)

2.    Provide the information students need to know in order to have an in-depth conversation about the subject.  Obviously, this usually will be a textbook or an article.  If there isn't anything appropriate, I have found it helpful to create it.  Rather than lecturing to fill in information gaps, I now write out the lecture, send it out electronically, and then spend the class time discussing portions that they don't understand and/or pushing the depth of their analysis.

3.    Define central goal(s) for that class. This needs to be done in conjunction with the two previous items and as you decide how you will approach the conversation.  Ask yourself, what do I most want them to understand by end of period and/or what is critical in this particular assignment?  If you keep your focus on the big issue as you plan, you will be able to include other bits as supporting pieces without obscuring the key idea.

4.    Create an open-ended, debatable question or hypothesis based on the central idea you want to cover.  This is always the hardest thing for me, but if you are really going to have a conversation it has to be something big and debatable.
As faculty, we have a real tendency to do the opposite.  Several studies have shown that 60-70% of the questions asked in sample of classrooms were memory questions ("Who was the president in 1990 or what is the definition of a molecule?")   Usually, only 15-20% of our questions required any higher order thinking. (The other 19% were "routine administrative questions.")

For me, key is starting with the central concept for the day and then working back to the supporting parts.  For example, when we are discussing the progress of the civil rights in the 1950s, my central goal is to get them to see that there were multiple factors at work, so I start with an overstated hypothesis like "Without the Supreme Court, African Americans would not have made any progress during the 1950s," and ask them if they agree or disagree.  This leads to the Brown decision, but it can work to black activism, Congressional action, migration, and other factors.  Forcing them to evaluate/assess like this requires higher order thinking as well as knowledge of the basic information.  On the other hand, if I ask "what was the impact of the Brown decision," most of what I get is simple memory regurgitation of what they read.

5.     Decide what role you are going to play in the actual conversation.  Do you want to be in the center, or on the periphery?   Having the faculty on the periphery means student led discussions, students talking to each other, and the professor limited to occasional interjections.  Obviously there is a continuum here, but at the other end, the faculty member directs everything and all conversation is channeled through the authority figure.  Student centered discussions can be real conversations and promote student responsibility, but if the faculty member is on the periphery, the conversation may wander and it is hard to push the thinking to a higher level.  Faculty directed conversations will always feel somewhat artificial, but are much more likely to discuss the central issues in depth.
I don't think we have to follow the same approach every day, but if we change our role in the classroom, we need to be explicit about what we are doing and give them time to adjust. 

6.    Plan the actual structure you will use for the day or for a series of sessions.  Naturally, there are lots of options, with some fitting particular topics better than others.
The classic approach is engaging everyone in a large group discussion on a topic you have asked them to think about.  The primary drawback to this is that participation is usually uneven.  You can try to counter this by direct questioning, asking for the views of a substantial number of students before evaluating/responding to the comments, or by having people sign in on the board with their response and then pushing the quieter ones to explain their brief but already public answer.
Lots of us start with small groups, having them work on the big question you have posed for the day, taking on a character in a role playing exercise, or developing a position for a debate.  Small groups force everyone to participate and offer the timid a place to try out ideas.  My preference when I use small groups is to halt them before the students feel they have exhausted the topic, move to large group, and push the depth of the discussion there as the groups interact.
Another version of the small group to large group process is to have them do concept mapping of a difficult idea or simply draw their understanding of it. (I provide big sheets of newsprint and have groups draw the Cold War. Then we process the results, looking for commonalities and missing pieces, and end by developing a composite concept map on the board.

Student led discussions are another great alternative. These can include small or large group work.  Two or three student leaders seem optimum, and you need to have careful discussions with them in advance on the central issues and how they will proceed.  Also, you will need to find ways of joining the conversation without taking over. The great advantage of this approach is that students feel more responsible to be prepared and join in.

Story-boarding can also work well with some topics, particularly if you have a number of separate but related questions or cases you want to discuss.  The idea is to have students move from site to site, where each site has a new question.  At each, they are to post their responses (via flip-charts, sheets of paper taped to walls, or post-it notes).  Once they have made the circuit, then you can work the results, summarizing, analyzing the answers and pushing the class to correct and/or fill in the missing portions.
All of these are basically variations on a theme, but remember, if you can create the illusion of variety in your classroom, your students inevitably will be happier campers.

7.    Prepare a conclusion of the main points or a summary of the discussion. This is a way of helping students see what was important, or what you want them to remember out of the conversation. I find it really helpful to write a summary of what we covered in the discussion before it happens - based on what I hope will occur.   This forces me to sharply define the key points and keeps me focused during discussion as well as helping me reinforce the key points at the end.
Some faculty assign a student or students to summarize the day's key points.  Another approach is to take the last couple of minutes to have the entire class to write down what they saw as the key points or their key remaining question.  People who use this tactic then start the next class with the results from that brief writing assignment.

8.    Learn your students' names.  Without this, it can't be a real conversation, and there certainly can't be any real accountability. 

9.    Develop system for reminding yourself who has talked and who hasn't.  Without this, you won't be able to push for a more inclusive conversation, and you won't be able to evaluate their class participation fairly.  Some faculty make notes during class.  I find that this distracts me and can have a chilling effect as students watch to see if you are going to note what they said.  My preference is to go back to the office each day and immediately go through my roster, making notes on quality and quantity of contributions.

10. Create an appropriate physical setting. If you leave the chairs in the classic lecture arrangement, it is going to feel like a lecture even if you say it is discussion.  At best students will see it as question and answer with the professor, and they certainly aren't likely to respond to another student if they are staring at the back of the speaker's head.  Circles, horseshoe shapes, double v's all work.  Be creative but do your best to rearrange things so that they can see each other (and the board, if you use it).

Preparing Students

For years I thought that if I could just come up with better questions, students would respond and we'd have great conversations.  Finally it dawned on me that I was proceeding on a huge - and largely fallacious - assumption.  I was assuming that they knew how to discuss.  Yes, they've had lots of practice talking with friends, but that wasn't exactly the model I was hoping for.  The kind of reasoned, evidence driven exploration I wanted was something that most of them had never experienced, so how could I just expect them to do it?.

As a result, I think we can help improve the quality of discussion if we spend some time preparing our students for their role.

1.    The first step is to discuss discussion. Talk about why you teach that way and examine the key components of a good discussion (it works really well to have them come up with these).    Students often find it helpful if you lay out the ground rules, such as whether they have to raise their hands and/or if everyone has to have a turn before a person gets a second turn.  (I generally find such things very chilling, so I focus on things like respect and civility.) 

2.    Create a comfortable atmosphere.  People are much more likely to take risks and to talk to each other if feel at ease.  This seems especially true for our students.  If they are worried about looking like a fool, they will keep their mouths shut.  In most classes I'm not willing to give up enough time to do ice breaking exercises with the whole class, but I find it very helpful to encourage small groups to get to know each other.  I therefore build in a little time for this in the first week or so of the semester, and then again when I change small groups. 

3.    Clarify your expectations.  They very often have a very different view of what they need to do than we do, so must be very explicit, and repeat often.  Explain how often you expect them to participate and what you value in their contributions.

4.    Make evaluation explicit.  For years I simply set aside a small percentage of the course grade for discussion and added it in when I made up the final grades, so students never knew what I thought of their contributions.  I gradually changed that, but the real breakthrough came a number of years ago when I started using a little form that I have them fill out on their participation at the end of each unit.  They turn it in, I comment on their self-assessment, and give them a discussion grade to date.  This process grabs their attention, makes my expectations much clearer, and really improves participation.

5.    Help students learn how to learn.  We need to explain how to read in the discipline and give them guidance on what to read for.  Suggest that they look for challenging ideas and/or things they would like to discuss.  Or that they look for the thesis, main points, and key evidence.  Explain how to mark up a book or take notes.  Encourage them to put the day's reading in the context of what they have already learned in the class.

6.    Provide assignments that push preparation for discussion.  These can be short pre-discussion papers, or simply a question to think about as they are doing the reading.  When I do the latter, I frequently start class by asking several quiet people to tell us their answer to the question.  You can also insure a higher level of effort by occasionally starting class with an in-class writing assignment on the question.  You can get similar results by doing a group quiz.  Another option is to have students put their names on the board along with their very brief answer (I often ask them to evaluate a policy and assign a letter grade) and then push the discussion from there.

Many faculty have found that discussion is facilitated by requiring an electronic conversation before class, and then using the class time to pick up on key issues from that preliminary work.

Another option is to have the students individually identify the key questions that need to be discussed, put them on the board when they come in, and then have the class reach a consensus on what it is most important to tackle first.

Doing it
Opening Gambits

Unless you are going to have students decide which of their questions to address, start with the question you told them to think about as they prepared for class.  Anything else is going to seem confusing or deceptive.

Ask the one question, don't rephrase, wait for someone to answer.  Study after study has shown that while we think we are being helpful by rephrasing questions, we are actually confusing our students who hear each rephrasing as a new question.  For variety, or if you can't get an answer after a reasonable amount of time, have the entire class write down an answer to the question.  This increases engagement, improves focus, and encourages the hesitant.

Let's say you are going to plunge right in with large group.  Depending on your central question, here are three possible approaches.

a)   "collective deconstruction"   Ask the class to generate list of key arguments/points they saw in their preparation for class.  Generate as many as they can and put them on the board.  Then (starting with the ones you see as central) ask the students about validity, evidence and assumptions until you have worked through whatever it is that you wanted them to see.

b)  "brainstorming a solution"   If you have posed a problem based question - a case study or why a complex chemical reaction works as it does - this is a fun approach.  Get several students to suggest possible causes/solutions/reasons.  Put them all up on the board without comment.  Have the class look for commonalities and cluster.  Work through the various solutions to evaluate and link back to the text or to prior knowledge.

c)  "hypothetical"   Start with a hypothetical situation based on whatever material you are covering for the day (clearly this works better in some areas than others).  Then ask, "If you had to deal with this, what would you do?"  I use this frequently to get shy kids started because there isn't a clear right or wrong answer.  Once you get three or four answers, then you can begin to push the respondents for reasons for their choices, and then bring in the rest of the class, linking the conversation back to prior knowledge or to evidence in the readings for the day.

Regardless of the specific approach, the key for me is to keep the opening free ranging and non-judgmental.  People are much more willing to get engaged if it is clear to them that the professor isn't looking for a single right answer.  I try to gather a number of brief contributions, and put them on the board (so we all remember their essence).  Then I pick and choose, developing those comments that I see as critical either in terms of my goals for the class or for explaining some area that many in the class don't understand.         

Keeping it going

1.    I think it is essential that we remember the difference between discussion and an oral quiz.  The goal of discussion is to get them to think more deeply about the subject and achieve a greater understanding of it; it is not to find out if they remember specific details in the readings. 
It is therefore essential to avoid creating the impression that we are looking for some pre-determined correct answer.  If we ask a question, get the right answer, and move on to the next question, we shoot ourselves in the foot.  Instead, we need ask broader questions and push them to explain why they answered the way they did.  Make it clear that you are looking for their views and their reasoning.

2.    As the conversation unfolds, stick to your announced role.  If it is student led discussion, fight the urge to take over.  If you are following more of a faculty driven model, make sure that you use your power to enhance the conversation.  Deepen their thought by asking follow up questions ("What led you to that conclusion?" or "You are absolutely right about that, but what about x?").   Push engagement with explicit pairings ("Mary, does Bob's point fit with what you were saying earlier?") and by calling on people ("Sam, Martha, John, we haven't heard from you today.  What do you think about this?).

3.    Monitor your actual behavior (as opposed to what you think you do).  I believe that there are real advantages to a more faculty centered discussion, but this can quickly lead to a monologue.  Several studies have shown that faculty who say they are doing discussion actually do 70-80% of the talking in the classroom.  If that's the case, then the students are not working and probably aren't thinking very much.

4.    Make sure that your questions are framed in ways that demand higher order thinking.  For example, see the handout of the Bloom and Ascher-Gallagher classifications from Diane Halpern (pages 96-7).  The key point is that real conversations and real learning are much more likely as we move to the higher, more complex end.

5.    Avoid counterproductive approaches.  Research shows that questions that don't work generally fall into three categories.  Some are too broad or vague, so students aren't sure what is being asked ("What about Ronald Reagan?"). The second category consists of questions that contain a series of sub-questions.  If they are all asked together, students become confused.  The third kind of question that often fails is one that is seen as having one right answer (yes/no or a specific piece of data).  Instead of asking "Is sulfur a pollutant," ask "Why is sulfur considered a pollutant."  Or better yet, expand it by asking "Why is sulfur considered to be a more dangerous pollutant than x."

Don't ask leading questions.  If the authority figure in the classroom starts a question with "Don't you think that...," the conversation is dead.

Don't answer your own questions.  Lots of us do this as a rhetorical device, but unless we are very careful, we can harm the conversation by sending a mixed message.  I think we are better off if we convince students that they need to respond every time they hear us ask a question.

Try to avoid jargon and/or terms that are over their head unless the technical language is central to the class and you have prepared them.  Ease them in to the language if you can.  For example, instead of "What is the author's thesis," start with "What do you think is the main thing we are supposed to learn from this?" Once they have the hang of it, you can explain "thesis" and use the discipline appropriate language.

Keep the conversation focused on the announced topic and avoid getting tied up in fine distinctions that matter only to people with doctorates in the field.

Avoid the "will all stupid people please raise their hands" questions.   Our standard refrains of "Do you understand, do you have any questions" require people to admit they aren't getting it, and most students don't want to admit that in front of their peers.  If you want to find out if they have questions, use "What questions do you have?"  This implies that you expect questions rather than asking them to say they don't understand.   Rather than asking them to reveal areas of uncertainty, ask questions that push them to demonstrate their understanding.   For example, instead of "Does everybody see how I got this answer," ask, "Why did I substitute the value of delta in this equation?"

Treat them with respect.   If they give an answer or ask a question that doesn't make sense, respond with, "If I understand your argument, ...."   Or "I'm sorry, but I don't understand your question. Can you help me out?   Then check back after you answer - is that what you were asking?

6.    Reinforce what you want them to learn.   Students often complain about discussion-based classes because they say they don't know what comments were on target and which weren't.  We can help them out without destroying the flow of the conversation by putting key points/contributions on the board.  We can also do very brief summaries as we move from one segment of class to another.  And finally, we need to do end of class summaries that focus their attention on the key points covered that day.

We can reinforce key ideas and encourage students at same time by praising especially good contributions when they occur, or by referring back to earlier contributors by name (especially as we summarize). 


As I have already noted, I really believe in ending the class with a summary of the central points covered in the day's discussion. 

If there were things that were important but you didn't get to, you can quickly point them out as you do the summary.  If they are things that students can understand without further discussion, then I figure that this "notice" is sufficient and expect them to be able to draw on this material in the future.

I find myself doing a lot more record keeping than I did years ago.  I return to my office after class, and record the quantity and quality of student comments.  I also make notes that may affect how I start the next class.  This may be an idea that they were struggling with, or it may an observation about techniques, group dynamics, or who needs to be prodded.  I will very often start the next class with questions to students (or a small group) that was fairly quiet in the last class.

Dealing with Problems

1.    How to draw out the quiet ones? I believe that a key here is having clear expectations, repeating them, and acting on them.  I start by explaining the value of discussion and explain that my goal is to have everyone engaged via a comment or two every class.  I follow up on that by calling on people, starting from the very first day.  That way being called on is more often seen as my idiosyncrasy rather than as an effort to embarrass them.  I also try to soften the approach by calling on several quiet ones at once ("Faye, Bob, Ralph, Martha, what do you think about this?").  One of them will usually respond fairly quickly, which breaks the ice and gives the others time to think.  Make sure, however, that you get answers from all the named people.  After the first week or two I use my records to make sure that no one goes more than a class period without getting pulled in in some way.

Shy students are usually more willing to speak if they are representing their small group, if they have had a written assignment on the question, or if I ask a hypothetical question where the answer is not based directly on remembering/understanding the material.

We can also encourage shy students with non-verbal signs, such as standing near them, or looking directly at them and smiling when asking the question.

I find it very helpful to encourage and/or prod the quiet ones in my responses on their end of unit evaluation forms.

2.    How to handle monopolizers? This is a significant problem in lots of classrooms.  Several studies have shown that in classes with fewer than 40 students, four or five of them control 75% of the conversation.  Obviously, this discourages the rest and kills any real interaction.

One thing we can do is to avoid encouraging the monopolizers.  All of us have a tendency to be thankful that someone is willing to talk, so we quickly welcome their contributions.  We have to remember that our goal is to have a wider conversation, rather than just listening to a select group.

A first step is simply to avoid eye contact with the monopolizers.  Another is to verbally silence the talkers and widen the circle ("We've heard from Sue, Jim, Cindy and Sara today.  What do the rest of you think.") If a monopolizer gets the floor and simply rambles on, we have to be ready to cut them off in a polite manner. I try to interrupt and ask a summarizing question ("Are you asking/are you suggesting x?") If the issue is worthwhile, you can then turn to others for their opinion; if it isn't something you want to give time to, suggest it is an interesting question/issue that you would be happy to discuss outside of class (very rarely will they do this). As a last resort, I have occasionally gone part way through a class, and then declared that the monoplizers are mute until I release them. (Do it nicely.  "Andy, Amy, Kathy, Paul.  You've been carrying the conversation today and I think it is time the rest of the class did some work.  So until I release you, the four of you are mute.")

The end of unit evaluation forms offer a great place to discuss this issue, and to suggest alternatives.  Sometimes it is best to invite the student in to discuss his/her role in the class rather than trying to put it in writing.

3.    How to respond to bizarre answers or questions? When I get off the wall answers I do my best to be gentle because I don't want to kill the atmosphere.  I've had success with "that's very logical, but actually not historically correct."  A less discipline specific approach that works for me is "could you point out the segment of the reading that led you to that conclusion."

Most of us get questions occasionally on things we thought we just covered, and therefore have to fight off the temptation to respond with "Well, as I already said,..."   I try to remind myself that my goal is to enhance their learning, and since they clearly don't understand, I have more work to do.  I usually start with "Sorry, let me try to make that clearer," and then use different language and different examples to go over the point again.  Using different language is essential both to keep the attention of those who got it the first time, and to help the one who is struggling.

I usually find tangential questions interesting so I have to struggle not to answer them.  If they don't fit the central message I am trying to get across, I try to respond with a "that's a very interesting question, but since it isn't tightly linked to what we are doing today, can we take it up after class?"

Fairly often students will ask questions that are germane, but pertain to something that you want to talk about later in the class or course.  My reaction to these is to say something like "That's a great question.  I think we will be able to deal with it more effectively if we wait a little.  If you don't feel I have answered it (by the end of class or next time), make sure you ask again."   When you do get to the appropriate place, name the student ("As Sara asked earlier") or have them ask their question again.

4.    How to handle arguments?  If the argument is between two students in the class, it is sometimes sufficient simply remind people of the ground rules for discussion (which hopefully you established early on), including the need for evidence rooted in the class materials and the concept that we focus on ideas, not people.  In other words, you can disagree with a person's views, but you can't attack the person.   If more intervention is needed, you can diffuse the situation and make it help learning by bringing the rest of the group in. Ask "What is the core of the disagreement?  Are there areas of commonality?  What assumptions are being made on each side?  Is there evidence in the reading that we haven't examined that would help us?" If that doesn't work, suggest that the dispute isn't getting the class anywhere, and move on to a new topic.  If one or two students seem to be central to the problem, individual conferences are a good option.  Do something: if you don't act to restore civil dialogue, the class will be destroyed.

Students sometimes will enter into contentious disputes with faculty members.  Sometimes it is simply a matter of misunderstanding, so I try early on to make sure I am understanding the student's position ("As I hear you, you are saying x") and find out what it is based on ("Help me understand.  What in the readings led you to that conclusion?")

If students persist, either in a single class or over the semester, there are usually deeper, more complicated roots involved than simply disagreement on an intellectual point.  Since those complicated causes deserve careful individual unraveling, let me limit my suggestion here to an exit strategy.  I don't see any point in continuing this kind of argument in the classroom; we aren't going to convince the student, we run the risk of appearing defensive, overbearing or arrogant, and we lose the attention of the class. As soon as you feel the situation has reached an impasse, end it.  The best approach I know of is to say, "Clearly we disagree.  If you would like, I'd be happy to continue the discussion outside of class, but in the meantime, let's get our focus back on what we need to talk about today." Then ask a question that moves the class to a new topic. 

Primary references: 

Barbara Davis, Tools for Teaching
Diane Halpern, Changing College Classrooms
Bonwell/Eison, Active Learning
Don Finkel, Teaching With Your Mouth Shut