share interactive transcript request transcript/captions live captions
download
|
MyPlaylist
NARRATOR: This is a production of Cornell University.
MICHELE MOODY-ADAMS: I'm Michele Moody-Adams. I'm a professor of philosophy in the Arts College and director of Cornell's Program on ethics and Public Life. And I'm also vice provost for undergraduate education. And it's in that capacity that I am delighted to be able to welcome you to this afternoon's talk, "Science Education in the 21st Century" by Professor Carl Wieman.
Dr. Wieman's talk is the inaugural public event, marking the recent creation of Cornell's new Center for Teaching Excellence. So we're especially excited to have everyone here for this occasion. The talk is also an important reminder of Cornell's longstanding interests in science education, and in the special challenges thereof.
Some of you may remember, for instance, a fall 2006 lecture in which Professor Eric Mazur from Harvard University talked about his work, research-focused changes in his course, that led him to be a converted lecturer. He called that talk "Confessions of a Converted Lecturer."
Now, I know that this afternoon our audience includes not only faculty members from Cornell, but also Cornell graduate student teaching assistants, and as well as teachers from local high schools and possibly middle schools. And so, we're delighted to welcome such a diverse audience of scholars and teachers to this afternoon's talk.
In the coming weeks, the Center for Teaching Excellence, along with the new Teaching Excellence Institute in our College of Engineering, will be co-sponsoring some discussion luncheons to encourage some follow-up work that will draw on the implications of Dr. Wieman's talk today.
If you've not had a chance to do so, you can pick up one of these blue sheets that will describe the events. One will be a luncheon for faculty members that will take place on Tuesday, September 30, 11:45 to 1:30. And the other will be a special event for graduate student teaching assistants that will take place on October 2, from 12:00 noon to 1:30. Again, there are more details on this handout.
So now, please welcome Peter Lepage, professor of physics, and the Harold Tanner dean of the College of Arts and Sciences, who will introduce our speaker this afternoon. Thank you.
PETER LEPAGE: Thank you.
[APPLAUSE]
I keep this introduction very brief. Carl Wieman is an extremely distinguished physicist. Winner of the 2001 Nobel Prize in Physics for his work in creating Bose-Einstein condensates, which is a very strange form of matter that happens when you cool things down to extremely low temperatures, much lower than outer space, in fact. I was going to say that it was the coldest place in the universe, except there could be a smarter Carl Wieman on another planet that made it even colder when he did it.
But we're not here to hear Carl in his capacity as Nobel Prize winner in physics, but rather because he has for many years been heavily involved in work, research, and innovations in teaching physics and now science more broadly to a broad range of students.
And he's created a Physics Education Technology program that created educational, online, interactive simulations, but also studied how effective they were in actually educating people. He's done research on student beliefs about physics, problem-solving skills. He's the recipient of the National Science Foundation's Distinguished Teaching Scholar award in 2001, the Carnegie Foundation's US University Professor of the Year Award in 2004, and the American Association of Physics Teachers' Oersted Medal in 2007. He's a member of the National Academy of Sciences and chairs the Academy board on science education.
I was particularly interested in having Carl come because he heads two major initiatives in science teaching, science education initiatives. One at Colorado, University of Colorado. One at UBC. These are initiatives that are spending literally millions of dollars on the improvement of science curricula, science teaching. And for me, I was very interested in understanding how one productively spends millions of dollars on this subject. And it's very rewarding to go and visit their website and actually get a sense of how they do it.
So with no further comment, I give you Carl Wieman.
[APPLAUSE]
CARL WIEMAN: Thank you. What Peter didn't say is we were graduate classmates together and shared a house together. And I was always very annoyed during that period how I would be slaving away doing these long problem sets, and he would sort of start the night before. We'd all been working for three days, and he'd-- oh, gosh, 10 o'clock at night. I'd better get started. And then he'd whip it off in a few hours. But I'm very relieved, hearing him talk about his work here that being dean, he's finally having to start to work for a living.
[LAUGHTER]
Anyway, I'm going to talk about really taking a scientific approach to teaching science. And I think what I have to say actually applies to many other subjects, as well. And as you'll see, I'll try and link to these ideas to basic understanding of how people learn. But I'm going to stick to the context of science, because, a, that's what I know the best, and, b, that's where we've got the most data, really demonstrating these ideas.
And I also want to apologize at the beginning, because there's a kind of delicate balance here that's very hard to optimize between trying to keep things as general, so one can see how there's fundamental ideas that apply in many different courses and, you know, educational levels that would be so relevant to the diverse groups of people I have here, and yet trying to give you specific enough examples to really understand how these really play out. And there's no real perfect solution to this. So I apologize ahead of time, if you can understand sort of the tension.
Now, before I start though, I want to make sure that you don't think you should pay attention to this because I have a Nobel Prize. It turns out, you get a Nobel Prize, you become an expert on everything, including education, no matter how little you actually know about the subject. That's for doing physics research.
The reason you should pay attention to this is that everything I say here is backed up by lots of data. This comes from researchers in actually a whole variety of fields, as you'll see, from all over the world. But a little bit of what I'm going to talk about actually comes from my own science education research group that I've had now for 12 or 13 years. And you can see lots of people who've been working in that.
So what I'm going to cover today is first briefly talk about why science education's important today, and then what research tells us about expert thinking in science, and the effectiveness of different teaching approaches at accomplishing that. And then I'll move on to spend a little bit of time talking about how one actually implements these research-based principles of learning in actual teaching. And then just spend a moment at the end mentioning-- as Peter talked about-- these science education initiatives that I'm leading.
So historically, science education was really about training this tiny fraction of the population that was going to go on to become the next generation of scientists and engineers. But over the past half century or so, there's been some big societal changes that have really changed the needs and purposes of science education.
And there's really two of them. The first is that we've come-- we, mankind-- has come facing these enormously important global-scale issues, which are fundamentally technical at their heart. And so we have to have a technically literate population if they're going to make wise decisions on some very critical questions, involving things like global warming and environmental change and genetic modification and so on.
And then the second reason is the economy. The modern economy is so based on science and technology that, really, pretty much independent of occupation, a person will be more successful if they have some basic technical literacy and complex problem-solving skills.
And so, because of these changes, what it means is that we, of course, can't neglect the next generations of scientists, but we also really have to think about making science education effective and relevant for a large fraction of the entire population. And so, that's really a much larger and in some ways more important purpose.
Now, I'm sure that it's going to occur to you through this talk of, OK, are we missing-- going to have to make trade-offs there between the training scientists versus this broader population. And I can talk more details in the question period, if you want. But all the research is showing us that there's no trade-off at all. If we do a better job at the one goal, we'll also be better at the other. We'll have more and better scientists, as well.
OK. So I've said we need to have an effective education for a large fraction of the population. What do I mean by effective here? And what I would argue an effective education is really changing how students think. So they come into our classrooms like this, and we want to have them go out transformed. See? That was easy.
Now, realistically, we don't really expect them to all turn into scientists like these two. But I would say that the basic goal in every course and over a program should be to have them in each step of the way, students come to think about and use science more like a scientist does. And that's really the measure I'm going to use when I look at how well we're doing, and how we might improve.
And having them think about science like a scientist is it really possible to accomplish this for most students, like I'm saying we need? And what I and maybe others are arguing is that it is possible, but it really requires us to take a different approach to science education. We have to start thinking of the teaching of science very much like science itself.
And to illustrate what I mean, I'm going to pick four basic tools or practices that are really at the heart of any experimental science program, and a big part, I'd argue, of why science is advancing so rapidly and successfully over the last couple of centuries.
The first is guide what one does by fundamental principles that come out of research in the context of education. That means principles of how people learn. Take the practices of what we do, based on good data and standards of evidence, not tradition, and anecdote, or even sometimes superstition, as is too often used in science teaching.
And third, is disseminate. After you've measured what works, disseminate the results in a scholarly fashion. And so as one's coming into a new area, you can learn about and copy what works and continually build from there, rather than reinventing each time some new person's coming in to teach something. And then, of course, fully utilizing modern technology.
And as they say, these are completely taken for granted in scientific research. And I'd argue that we really ought to have them being in science teaching in the same way. And so, the rest of my talk is really going to try and support this way of thinking about teaching science.
So before I get into a bunch of research studies and results, I want to frame this a little bit by giving a perspective of how I came to think about teaching science in a very different way. And it goes back to, OK, how'd I start?
Well, when I was first called upon to teach physics, many decades ago, I used what I think is kind of a basic human reaction to ever teaching anything, which is that I'd go off and think. I thought about this subject-- I think it was introductory electricity-- really hard, so I could get it figured out in my own mind really clearly, so that I could then go and explain it to the students so they could understand with the same clarity i had.
Well, that was the idea anyway. But I've always been a really hardcore experimentalist, and so I would measure what they were really getting out of this. And so was clear that my brilliantly clear explanations were leaving most of the students really quite baffled. And when I measured, somewhat subtly, some of my colleagues, the students, they really weren't doing much better. And so for a long time, this was just a frustrating puzzle to me.
But the way I came to make some progress was actually by looking at my graduate students. I had a quite big successful atomic physics program for many, many years. So I spent a lot of time paying attention to the development of my graduate students, and seeing how they progressed.
And I came to see a very consistent pattern that emerged. And that is that students would come in to work in my lab. And by definition, they had 17 years of great success in science classes. Otherwise they wouldn't have been coming to work in my lab. But when they'd start to work in research, they were really, basically, pretty clueless about how to actually do physics.
But-- and here's really kind of the surprising thing-- student after student after student, though, after just a few years of working in the lab, they were suddenly expert physicists. They were winning arguments with me all the time.
Now, the first few you see this, you think, that's just curious that that student is like that. But as I came to see this happening over and over, I realized, no, there's something more basic going on. This isn't a characteristic of an individual. There's some fundamental pattern here. And I really wanted to try and figure out, get the bottom, could you make sense of this?
Now, one hypothesis that occurred to me-- and I think it's occurred to many others who've seen this same thing happen-- is, well, maybe the human brain just has to go through sort of a 17-year caterpillar stage before it can blossom into a physicist butterfly. But I took a little more rigorous approach to this and really started to systematically look at research on how people learn, and particularly in science, to see if I could understand this pattern.
And what I came to realize is that, OK, caterpillar to butterfly wasn't the explanation. But then, in fact, this pattern and change here did make sense. And in understanding it, it also to convinced me that there was tremendous opportunity to improve the learning by students in regular courses.
And this was really based upon looking at the research, and particularly the major advances in the past decade or two. Across these three rather different areas, the breadth of cognitive psychology, to basic brain research, to classroom studies, particularly in the college and university science classes, there's a very consistent picture between all this research that told us really what was important about achieving learning. And so, I'm going to try and give you a few samples of that, how I came to think about this in a very different way, and how to think about teaching in a very different way.
And so, I'm going to start by looking at some examples of research, first on the ideas about what's been seen about how experts think and learn, and then research on traditional science teaching, and how well it's teaching expert thinking, and then some on research showing us how to do better.
OK. So it turns out cognitive psychologists have done a lot of studies of expertise. And they find that over of very wide range of disciplines, academic disciplines, and even actually some athletic activities, playing chess, et cetera, there's a few very common, consistent characteristics of experts. The first one's no great surprise. Experts have a lot of factual knowledge in their area of expertise. But the others aren't so obvious.
It turns out that experts have a unique mental organizational framework for the information that's consistent across that discipline that allows them to very effectively retrieve and apply their knowledge. And this is based on certain kind of expert recognition of patterns, associations, connections, et cetera, which in the scientific domain usually-- part of that, in scientific domain, is really what we think about as scientific concepts. That's a way that experts have taken vast amounts of different kinds of information and fit it together in a broad framework, that allows them to use that information and apply it in certain ways very effectively.
Now, a third characteristic of experts is they have an ability to monitor their own thinking and learning, at least in their discipline of expertise. And so they're able to ask themselves, do I understand this? Does this make sense? Am I learning this right? How can I check that? And I think many teachers, actually, who've developed this-- and certainly it was true in my case-- just assume that's something that people do, and it isn't. OK.
Both this organizational frameworks and this monitoring thinking are very much acquired expertise. And I think implicit in traditional science teaching is the idea that, well, we give them the knowledge, and they already have these other things or is it kind of comes along automatically.
But that's really not what the cognitive psychology tells us. That work says that these things are developed. They require many hours of intense effort and practice to develop them, with appropriate guidance and reflection. And in fact, to reach a high level of expertise in pretty much anything, takes many thousands of hours of intense effort. And sorry for students who find that depressing to know, but that's the reality.
And I think you can make a pretty good case that this is fundamentally pretty biological in origin, that everything we think about as expertise is really stuff that's in the long-term memory. And to develop the long-term memory, your brain has to build little proteins and put them and the right structure in the way. And that just is a slow process. It takes a long time, and there's just no basic shortcuts for it. But you know, what it means is, OK, these things have to be developed if they're going to be acquired.
So let me turn now to say, OK, I've argued that we know certain characteristics of expert thinking. We'll look at how well students are actually learning that from traditional science teaching. And by traditional science teaching, I mean the kind of teaching that's, I think, likely familiar to most all of you who've had a science class. A great majority of science classes work this way, where the primary contact between student and instructor is the instructor standing up like this, lecturing to a largely passive, silent group of students. And then those students, they go home, and they do back of the chapter problems from the textbook, and they have exams are similar.
And so over the past 20 years, people have been doing a bunch of research on, OK, how well's that working? And I'm going to focus on two aspects of this. One is conceptual, understanding mastery, and two is just basic beliefs about physics and equivalently chemistry, and really what it is and how to learn it-- how well students are acquiring that from their courses.
Now, it turns out in physics, I think probably because we physicists pride ourselves that we've got a few basic concepts that are widely applicable. And so physicists have done a lot of research on how well are students really mastering these concepts, particularly in our introductory courses. And in this process of studying this, they developed a number of assessment tools that are quite carefully tested and validated and so on, that really are quite accurate at measuring the student mastery. And one of the oldest and most widely used is something called the Force Concepts Inventory.
Out of curiosity, how many people here have heard of the Force Concept Inventory? OK. So only a small fraction.
So what this is is it's something that-- it takes a subset of the basic concepts of force and motion that are covered in pretty much every first semester physics course taught anywhere. And it tests students' understanding of these, normally by asking them to apply them to some simple, real-world situation, like car running into a truck, or in the question I display here. This is a hoop sitting on a tabletop, and a ball's rolling around inside it. And what direction does it go when it flies out the end there?
And the way this is used is then it's given to students, typically, at the start and then at the end of their course that covers this material. And one looks at, OK, what percentage of the questions did they get wrong, not know the answer to at the beginning, that they actually got right at the end? That tells you what percentage that they learned of what they didn't know. And this is now pretty routinely given in hundreds of physics courses every year across the world, to measure this learning gain.
Now, what's emerged from this is a really kind of remarkable result, which is-- and I displayed it with this histogram. But what this is looking at is you take that average over the whole class, and you look at what the average amount the student learned, this learning gain.
And what one sees-- and this is a histogram showing the results from actually this average class from 16 different classes. What you can see here is that in the traditional lecture course, the average student never learns as much as 30% of the concepts that they didn't know when they started the class.
And what you see in this data-- and I have piles of other data that's unpublished, and there because usually the faculty do worse than what's shown up here and don't want to it seen in print. But this result shows up independent of lecture quality, class size, institution. In fact, I think in here is Harvard University and a couple of community colleges. And you can see, they're all in the same boat.
It says it that this kind of teaching is simply not effective for people learning, developing a basic mastery of concepts. We now have data from a variety of other levels, and we have data, similar kinds of data, coming in from other disciplines that are giving us the same consistent picture. So this is saying expert organization, in terms of concepts, is not coming through in the traditional approach to teaching science for most students.
I will say that partly driven by this, as well as other things, people then have worked on different ways to actually teach introductory physics. And I'll talk a little about those later. But they're now pretty consistently coming in with factors of two to three higher learning gains than with the traditional approach.
But before I get to that, I want to turn to a very different aspect of learning. It really doesn't involve the content so much at all. It's this basic beliefs about the subject. And it turns out, if you interview lots of people on their beliefs about, say, physics and how it's learned, you find that there lies on a novice to expert scale, where the characteristics of novices and experts are, the novice sees the content of physics as just isolated little pieces of information, and you learn physics just by memorizing all these little pieces. This information is just handed down by some arbitrary authority, and quite disconnected from the world outside the classroom. And novice problem solving is matching the patterns of the surface features of the problem to certain memorized recipes. OK. And most physics professors will be nodding their heads that, yeah, they've seen all this.
Now experts, like a practicing physicist, they see the physics as this coherent structure of these very broadly applicable concepts. These concepts describe nature and are established through experiment. And expert problem solving is using these systematic concept-based strategies that looks at much deeper structure and patterns in the situation, which makes them widely applicable, including into completely new situations and contexts, and therefore are much more useful than the novice-like approach.
So once you understand people's beliefs lie on this, you can develop surveys that actually probe, that it can actually measure them. And this is one thing my own group has done. And probably the most widely used survey now is on this is one that we've developed and tested. And so, you could now use this survey just like the Force Concepts Inventory. You can measure your class, what their beliefs are at the beginning of the term, and then you can see, OK, how much more expert-like have they become as a result of taking this introductory physics course?
Well, that may be what you'd like to be measuring. The reality of what we and others who have used this measure now in many, many courses is virtually all introductory physics courses, actually-- except for a couple experimental ones I'll tell you about-- leave this students significantly more novice-like in their beliefs about physics as a result of taking this course than they were before they ever started.
Now, if any of you who are laughing are chemists, I'll point out that we recently developed a comparable survey for chemistry. And we see, if anything, chemistry's as bad or maybe even worse.
OK. So that's a sort of different aspect of expert-like thinking. But it's telling you know that there's a real problem here with traditional science teaching, in terms of what we're accomplishing. And so, I want to say, look at why is this the case. And can we dig a little deeper here to understand this?
And I think one part of this problem is that the lectures are presenting information. They're presenting ways to solve problems, and so on. And to them, they're expert instructors. But the relevance and the conceptual underpinnings that are so clear to them, are so sort of ingrained, that they just can't imagine it's not being obvious to everybody.
And I talk about this that psychologists talk about, the curse of knowledge of how incredibly difficult it actually is, you've know something and mastered it, to understand how somebody doesn't know, doesn't know anything about it, how they can see that same situation and information. And so, that's just an extremely difficult thing to do.
And so the novice-- and presumably most of our students in our courses are pretty novice in the subject. Otherwise, why would they be taking a class on it? That same presentation to them can seem very differently. It can actually reinforce their perception that, you know, all this is about a whole bunch of memorizing these facts and recipes. And that they not only never have the opportunity or never practice this expert-like thinking, they never even are aware there is such a thing as this kind of different expert-like thinking and problem-solving.
And then, a second aspect is the one that's kind of much more simple and basic. But it's why there's a real problem with what happens in a typical lecture class. It's the limitations on the working memory that are not being properly considered here.
And what this is, what I'm referring to here, is something that actually is a basic element of how the mind works. And it's really nicely represented by this Gary Larson cartoon. This is an extremely well established result from cognitive science. And what it says is that the part of the memory that deals with remembering and processing information on short time scales, like would be relevant to an hour-long science lecture, has extremely limited capacity that's in contrast to the long-term memory, which has enormous capacity.
But this working memory capacity, it's, in fact, has people say this very carefully. And there's a bit of an argument now as to whether the typical human brain can deal with a maximum of seven distinct new items, or only four distinct new items. But however you slice it, that's just a tiny amount compared to what a student is exposed to in almost any science lecture.
And it's also true that the way that working memory functions is the more new stuff it gets, the more it kind of slows down and has trouble doing anything. It's a whole lot like a computer with not nearly enough RAM. But it's only has to get up to something between four and seven, and that's when you get the screen of death that just nothing more happens.
It's hard to really, until you put it in this perspective, to think just how remarkable it is. I mean, that means that some new technical term that a student hasn't heard before, even if it gets explained to him, that's just used up something like 20% of working memory that they have for that whole class period. So it's really pretty dramatic here.
I say, since all of you have human brains, and I'm pretty confident they work much like this, and I'm going to be pushing your limits here, I will make sure my PowerPoint slides are available, so you go back on where I exceeded your working memory and review. Sort of artificial long-term memory, here.
But what this means is that the amount of processing and retention from the typical lecture is really tiny, just because of these basic limitations. I want to emphasize that's very much the case just for novices. This is a very expertise-dependent thing. And it's because if you know more about the subject, then the stuff they're talking about is more calling on stuff that's in your long-term memory, and you don't need to use up short-term memory for it. And frankly, I'm assuming in this audience that most people here have thought about or are familiar with many of the general issues I'm talking about. Otherwise, I know that once I'm 1/3 of way through, it's kind of gibberish beyond that. But hopefully, many of you are in that category.
Anyway, it's really easy to confirm this. And I'm going to give you a couple of examples of two very simple studies that anybody, any instructor who had a strong stomach, could repeat themselves. The first is from Joe Redish, who's a physics professor at University of Maryland and is actually, was and is, considered one of their outstanding lectures. But he's also a very thoughtful guy, and he knows about physics education research. In fact, he's in that field now.
And so he started really worrying about measuring what his students were learning. And so he just hired a graduate student to stand outside the door and at the end of the lecture, randomly grab students and then interview them, ask them, what was the lecture you just heard about? And what he found is, consistently, the students could give only the vaguest of generalities in response to that question.
In a little steady that Kathy Perkins and I did, we would take examples out of the middle of the lecture of some significant but non-obvious fact presented in lecture, and then test them 15 minutes later. What fraction actually remembered that? And we typically got numbers like about 10%.
And there's many, many other studies that show you quite comparable things about how little a novice in a subject can really get out of listening to an hour-long lecture, unless there's a whole bunch of very special things done. So I'm just taking kind of the traditional situation. There are ways you can make lectures effective. I'm just looking at the typical case here.
OK. So getting back to the puzzle that got me into this in the first place with graduate students. You can now see how all of this actually makes very good sense, that these students were doing great in these classes, but those classes were really not teaching them expert-like thinking in physics. But when they got into the research lab, in fact, they were doing exactly what we know is now necessary for developing expert-like thinking. They were very strenuously engaged in thinking about the physics of situations, thinking about the associations, the patterns, applying information continually, and getting guiding feedback to sort of coach them in that thinking.
And so it wasn't anything magical about the atmosphere of the research lab. It was simply the cognitive processes, that they were spending all their time in there, compared to the cognitive processes they'd been using in their classes. And that's really the explanation for the difference here. And it also says, gee, so why don't we get those cognitive processes into all the classrooms, and students will do a lot better. And that's exactly what lots of people, particularly in physics education, have been doing on that.
So I'm going to sort of turn to, OK, so how do you do that? And what really do you have to build in to make sure people learn? And I mean, it's a pretty straightforward process. Now, straightforward does not mean easy. There's a lot of work involved. But you can really argue, approach this in a very scientific way. We've got certain principles, and if you get those, you follow those, you'll get a great majority of really what matters here, from the first few principles I'm going to give you. And these are all coming out of lots of different people's research.
The first is learning is always built on prior thinking. You're changing a brain, and it starts out in a certain condition. To be effective teaching, you absolutely have to connect up appropriately with the students' prior thinking and understanding.
Secondly, there's got to be an explicit-- you know, they've got to see what expert thinking's like, and they've got to explicitly practice it for an extended period of time, and at very strenuous level. We know that's required for really developing this the level of expertise. In fact, people have stated the brain, in recent years, have come to recognize brain development's much more like a muscle than they used to think.
Everybody knows that if I want to build up a muscle, I got to go exert it strenuously, and I got to do that over and over for a long period of time. People now recognize the brain is developed in very much the same way. And so you've got to have that kind of strenuous, intense focus, therefore, in engagement and thinking about it.
And you've got to have effective feedback to guide. Just thinking about something doesn't work. You've got to be guided to do in an effective way. And we know that to be effective, feedback has to be timely, namely, right when they're thinking about it, and it's got to be specific. So it's quite specific to give them useful guidance on this.
But that's a whole lot of work. I talked about all this extended, strenuous effort. Human beings evolved a long time ago to have sense enough not to put in a whole lot of effort on something unless there's some clear reason to do that. And so that comes into motivation. In order to learn, they've got to be motivated to put in this kind of strenuous effort.
And then there's some littler things here. You've got to worry about these limitations on working memory and the details of what you work with. And then I'll point out one almost-- well, it's not trivial. It's just a clear, straightforward, but it's not often very used. OK, we also worry about retention. Basic information and processes, people know exactly how to have that retained.
And there's no secret about it at all. You have to have a person that has to go through spaced, repeated retrieval and application of the ideas and processes. And you've got to build connections to other things. You do that, it gets retained. You don't, doesn't get retained. Period. So pretty straightforward.
OK. So in the interest of time, rather than these other things I've grayed out now, are relatively straightforward to think how one can apply in a particular course or context. I'm going to talk a little bit about two and three here in a little more detail though, where it may not be so quite so obvious.
Motivation has been studied a lot. And a few basic things. First, you got to realize it's a pretty complicated subject. It depends on people's previous experience, culture, all kinds of stuff like that. But some things we do now-- and then this actually comes from some studies from my own group-- is it in terms of learning science, a key element in seeing that it's interesting-- and therefore that's a big part of motivation, is being interested to learn it-- is seeing relevance or usefulness.
We do a lot of surveying, and we've got now thousands and thousands of responses to students as to what made their increase in a particular area of science. What enhanced their interest? And it's overwhelmingly that they saw how this could be useful or relevant to something they didn't know about before. So that's a big part. And that means you really have to present things in meaningful contexts if people are going to see they're interesting.
There's an important, big aspect in motivation, is person has to have a sense that they can master a subject, and also beyond that, a better sense of the process by which they need to go through to master the subject. Both those are quite relevant to motivation.
And then a sense-- especially if you want somebody to work really hard at something-- a sense of sort of personal control, or some level of choices involved in it that they're responsible for. Also, this is clearly identified with motivation.
Now, I might say before I leave this, I run into a certain number of faculty-- and probably they're not in this room-- who argue that, gosh, by the time students come to college, I shouldn't be worrying about having to motive them, tell them this is interesting. They should know all that. That should have happened when they were 13 years old.
And the only thing I can say in response to that is, well, if it didn't happen when they were 13 years old, and it only happens when they're 13 years old, if they happen to be awfully lucky and to where, you know, what socioeconomic class they were raised, who their relatives were who talked to them about science or whatever, they didn't get that, then if you don't worry about motivation, you might as well just write them off.
Because there's nothing genetic about an interest in physics. It's developed. And if you want your students to learn it, you better convince them that they should be motivated. Why it's worth learning.
OK. But you motivated them now. And so what you motivate them to do is really practice expert-like thinking. And that means getting them engaged and then monitoring and guiding that thinking. And what we know about acquiring that is that people have to go through a series of ongoing more increasingly difficult tasks, where each one is challenging for them, but it's doable. And so these are task they're thinking about out working and completing.
And it's really important that one explicitly focus on these expert-like ways of thinking, which aren't obvious at all, if you're not an expert. And I've talked a lot about the idea of concepts, how things are related. That means in developing this, they have to be tasks that really call upon them to explore relationships and associations between different pieces of information, different phenomena, et cetera.
Working on sorting out what's relevant and irrelevant information in a certain situation for solving a problem. That's things experts are really good at that. But it's a very much acquired expertise. There's got to be a quite explicit focus on this kind of metacognitive processing of experts. The self-checking. The saying is, am I making sense here?
And then finally, the reflection on learning. Am I learning this in a sensible way. How does this fit in with the things I've learned before? How much do I have to modify them, et cetera.
All of these things are developed through practice, but have to be part of that practice. And you need to give students-- to learn, you've got a give them tasks that quite explicitly have them do this. And the data says that when they do this, they get a lot better at it.
OK. So I told you what you need to do. And I think most thoughtful instructors dealing with one or two, maybe three, students could probably do that reasonably well. But it's been a long time since I had a class of three students. And I've been classes of 200 plus students for some time. And so the question is, how do you do this with a typical, large class?
And what I would argue is actually more than about five students, you really have to look to technology. Here's a place where technology can be a big help. In fact, personally, I'd argue it's probably essential to be using technology to make this work.
And I'm going to talk about a couple of different technologies. Well, no, I'm not. The first I'm going to really talk about, just the highly interactive lecture, as supported by clickers. The other technology, if I'd had more time-- interactive simulations I think I'll skip because of time today.
So how many people here are familiar with using clickers? OK. So most of you, but not all. So a clicker is basically a personal response system. Often, they're about two or three times bigger than this laser pointer. Every student has one. And the way they work is that in my class, for example, I would ask these introductory physics students this question about, OK, this basic conceptual question about what happens to the brightness of the light bulb when the switch is closed.
The students' clickers have a series of buttons on them. Usually, I bring one to show off, but I forgot to bring it on this trip.
SPEAKER 1: There's one sitting on the table right over there.
CARL WIEMAN: Oh, perfect. In fact, it's even the kind I'm familiar with. Yeah, so it's got these buttons on it. The students think about this question. I might have them talk to a few of their neighbors before they answer. And then they'll push the button that they think corresponds to the answer they think's right. I've got a receiver on my computer. Because these things are coded, it picks up who that student was and what answer they chose. And then usually, after all the answers are in, it displays to the class a histogram of all the results.
Now, this is a technology that the initiatives I work with have been doing a lot of work on helping faculty use them, and a lot of studies about how they can and can't be used effectively. And so we've got a bunch of data on this, which we're busy slowly publishing. But in the meantime, we've written a pretty extensive guide on details about how to use these things most effective way that's on our website, if you want it.
But just a few brief things I'll say about this. First is, like all technologies, this isn't automatically helpful. One really has to think about what you want to accomplish as a teacher, and how this can support you in accomplishing that.
And we've seen many cases where they're basically just used as an expensive way to take attendance and test the students a lot. And the students are pretty grumpy about that, and probably it's not very educationally useful. They're grumpy particularly because they got to pay money for these things.
But when they're used and perceived to be used really to enhance this kind of engagement and communication and useful feedback, then they really can be quite transformative in a lecture hall. And some of the key elements we've seen to make this happen is partly just the attitude the instructor, but a few details.
First, it's important to have challenging questions. Oftentimes, questions are made too easy. Usually good to have them focusing on these deeper conceptual issues. Student-student discussion-- like peer instruction like Eric Mazur talked about when he was here last year, I'm sure-- is a really valuable part of this.
And it's valuable both in the learning that takes place in the discussion, but also the feedback. Students get a lot of feedback just from talking to each other that can be a lot more specific than anything you can do with a 200-person class. But you do, as an instructor, need to follow up on the discussion to be as most effective feedback as you can.
In fact, I've gone so far in my classes to actually have assigned discussion group. So each four people here will actually, on these questions, for certain questions, they've got to come to a consensus before they actually answer. So they're all busy discussing these questions.
What that's really useful-- as well when it's useful for them-- the way it's really useful for an instructor is you can wander around and listen to those discussions. And then you not only have the information about student thinking from seeing what that histogram is, but you've listened in. You know very specific things that they're having trouble with, questions, what they understand. And in the follow-up discussion, you can bring those out and therefore make this much more specific and therefore much more useful feedback to students afterwards.
Does that mean I'm running too long? Boy. This is a tough place here. What?
SPEAKER 2: The clicker did it.
CARL WIEMAN: The clicker did it? My. OK. So I better move along quickly here.
So you've done all this. Your classroom's incredibly interactive, and they're practicing expert thinking. It's not enough, OK. And I emphasize this because a lot of really thoughtful instructors put a tremendous amount of time thinking about preparing what's going on in class. They don't put much time at all in-- they just say do problems 5 through 15 at the back of the chapter. And that's really a mistake.
And the reason it's a mistake is back at this issue of basic biology. There's no way in the time that students have in your classroom, no matter what they're doing, to really develop long-term memory enough. They've got to have more time. And that means that you really have to send them out of the classroom to spend many more hours practicing this expert-like thinking. That means designing good, authentic homework problems that give them further practice in that, and find ways to get them effective feedback on what they do.
OK. So the interactive simulations we're skipping, because I'm running too late. OK. So I'll just start finishing up here. So I've been pushing all along, one needs can take this kind of scientific approach using basic principles to actually improve instruction. But it's also really important that we do rigorous measurements of how things are working.
So I'm they give you some data of what sort of things we and others see when putting those things into practice. So I'm going to just take the particular examples of where I already presented data to you.
The retention of information from lecture, which I don't really want to argue as a particularly important thing anyway, but we have just looking at how to make it work, we get enormous improvements in that when we work on it. More important things, like the conceptual understanding, as I already mentioned before, there's two to three times improvements now getting pretty routine.
And in these beliefs about physics and what it means to learn and solve problems in physics, this is a much newer area. And so it's only been a few years we've had the tools to measure it and start understanding what was relevant.
So in the very early minor interventions, we found that we could at least pretty easily get rid of these significant drops. But it's only been actually in this past summer I've been very happy to see reports come out of a couple of new experimental courses, where they didn't do drastic things to the content. But by really carefully understanding and tailoring it to address these student concepts, we now, for the first time, have two demonstration examples that it is possible to actually, within a single physics course, to make substantial improvements, in other words, to have students think about this learning physics in a much more expert-like way. So that's just more data showing this really can work.
OK. So just finish up here with talking about, OK, we've got these big great ideas. How do we actually make them widespread in every university classroom. And I think it's clear that it really requires changing the culture and the thinking about education, starting with the major research universities, like here. And you've got to work at the department level. And that's really what these initiatives that I'm leading are doing.
And they're going through and providing a bunch of resources and a bunch of incentives to have departments go in a systematic, scientific way to go through all their main undergraduate courses and figure out, OK, what are the real goals we want students to be able to do after successfully completing these courses? Let's develop good, rigorous, objective measures of how they're succeeding. And then putting into practice either tested best practices or developing new practices, guided by these principles of learning, and measure that they work in this kind of systematic way.
And I'm quite optimistic if we can get educators to really approach teaching science in this scientific way, the results will both be dramatic improvements in student learning, but also in the same way scientific research has become a very efficient process, because we don't have to copy things-- we learned so much from each other-- we'll also make teaching more efficient.
And in that spirit, we're making sure that all the materials, tools, data we develop are all going to be accessible on the web. As Peter says, we already have some things there.
OK. So I'll just end with this, that I think we need a much better approach to education, and this scientific approach provides it. And if you want to learn more about these ideas, just some good references, starting, this NAS book on how people learn. That's sort of the longest and most well-written, and they work their way down to the shortest and worst-written by me.
And then also, if you're interested, some more details on the research, the [? CLAS ?] site here actually is where my own group has the surveys and a lot of the research papers based on that.
So thank you. And I gather there's supposed to be a bunch of time for questions, arguments, discussion now. Right?
[APPLAUSE]
Sure. I'm happy to handle questions. If they start throwing things, you might need to intervene. Yeah.
SPEAKER 3: My wife is a speech therapist, and she describes encountering a recurring problem with many of her students that she relates as being ready to fire, oops, A. And I see the same kinds of things in my students, as well. And in part, I think this might derive from a sense that we have instilled, that they have to give an answer very quickly. And that's certainly reinforced by exams that have to be done in a limited amount of time. Where does this fit in at all?
CARL WIEMAN: Yeah. Yeah. So if you couldn't hear him in the back, he's wondering about the tendency for students to feel they have to give an immediate answer, and the way exams might reinforce this. So, yeah. So a couple of things I can say.
First, that's very much a novice idea about problem-solving, that namely, you have memorized answers. And so that means if you know it, it's something you respond immediately. That's the answer. That answering something, solving something is not a reasoning process that might take some time. So that's something that they've learned through bad teaching, essentially, or misinterpreted teaching, is really better than bad.
But one of the ways exams really aggravate this is that in the problem-solving research literature, we talk about exercises versus problems. A problem is something that you don't know how to get this. You don't know the solution. You can reason your way through to it using certain approaches, but you can't just give the answer.
An exercise is something that you've work 50 problems that look just like this. And so I give it to you, and your expert brain-- this is actually quite literally true-- your expert brain just goes right to this one little place where information's stored. It doesn't have to activate any of these broader control things. And that clicks out, here's the answer. And functional MRI will show that it activates just a little, tiny part of your brain doing that.
Now, when we give exams to students, most of the time you can only get through a typical university exam in time if you're doing those as exercises. In other words, you've practiced and practiced and practiced so many times that that's just an automatic thing. You don't actually have time to go through the reasoning process. And very few exams ask you to go through the reasoning process. And so exams actually really-- when we talk about them getting more novice-like-- you've just put your finger on some key elements of why that's happening. In back.
SPEAKER 4: You mentioned that the short-term memory can only accommodate four to seven items, but the totality of the items on the list of physics curriculum divided by the number of class dates is certainly more than that. So can you elaborate on exactly what one does about this?
CARL WIEMAN: Right. So it's vastly more. And particularly when you start including the number of technical terms alone, without any understanding beyond technical jargon, probably, it hits the limit. So one has to think very carefully about what you can do. You can't eliminate, you know, cognitive load, demands on long-term memory. But you can think very carefully about how to eliminate unnecessary cognitive load.
And so one thing is you just have to be realistic about what people can cover. A second thing is you have to recognize that, OK, if I want to cover these things, students ought to be able to learn something outside of class. And certainly, there's a whole bunch of information that they can learn.
Because what I talked about with working memory limits, those are within time-constrained values. If it's something that somebody can read the book and think about and work backwards and go through slowly, then you don't have those same constraints. So the best thing to do is say, OK, look, realistically this is what can be accomplished in a class. I either have to leave out some topics or have them learn them on their own, independently, which isn't such a bad thing anyway.
But things you can do, literally in the class, just think about everything that takes mental processing. So I'll give you a couple of quick examples. I already talked about jargon. Don't use it. Introduce it outside of class. They got to learn about it-- or if you want to use it, if you want to introduce some technical term in class, think about, gee, is that really 20% of what I want them to learn today? Usually, it's not.
But other examples that demand working memory are when there's a process or an image, where you just say it in words, and they've got to use their brain to think about what that shows. If you just give a simple visual sketch of it, it dramatically reduces the working memory demands.
If the organizational structure is very clear, and so they see how different things relate together, then that means those things can get chunked. That's the technical term. And so if they see this that is very intimately related to this, as the same idea, than that takes up half as much working memory as if you didn't take that extra step to make sure they saw how they all fit together, and so they saw them as separate ideas.
So those a few of the elements. Actually, I got a whole bunch of specifics in my slides that I never got to. But you talk about a big issue and small issues that come into this.
SPEAKER 5: You mentioned outside learning. How do you feel about pre-learning versus getting them to read things before they come to lecture versus after lecture and going back? Do you try to have pre-reading?
CARL WIEMAN: I require pre-reading. And it's just for the issues of working memory. If you think about what the unique thing an instructor can provide to a student, reading a textbook is not. They can actually do that without you standing up there and in one form or another reading it them.
And so what I would argue is the most efficient use of this precious resource, and expensive resource of professorial time, should involve the minimum of just giving them basic information. So you should think about the basic information they need in preparation for the class. And they should be expected to read that, maybe even tested briefly on it. But then in class time, they're spending all their time thinking about this, processing, and applying that information in these kind of expert-like tasks.
And so in the process of doing that, there's going to be many questions come up that get at the basic ideas. OK, they read it, but they didn't really understand it. But now you're fitting that information into a framework and a structure, and they see where it's needed. Then they absorb it much better. So you actually do end up transmitting a fair amount of information, but it's in a much more useful and effective way. Yes.
SPEAKER 6: Have you dealt much with student perceptions of teaching using these new techniques and approaches with group work and such? And, I guess, one reason I ask that is because so much of our teaching is constrained potentially by student evaluations. And so I was looking for you to address that.
CARL WIEMAN: Boy, am I glad you brought that up. If you couldn't hear that, she's worrying about student perceptions. And I put it in another term, student resentment, in fact. And so, yeah, we actually have looked a lot, a fair amount, about this. And I can tell you several things that are quite important.
So it's not at all unusual for there to be some student resentment. People automatically resent anything they're not used to. And so these students have been seeing this all along. Then they resent it.
Eric Mazur, who maybe some of you have heard before, he was just talking about how he's compared his student evaluations. And if he follows an instructor who's been using the very traditional lecture type, Eric's student evaluations are among the lowest in the department. If he follows an instructor who's been using this peer instruction and highly interactive lecture, his are among the highest in the department. So it just shows that these things are very preconditioned on what people expect and what they're used to.
But we've actually been I won't say universally successful, but quite successful, at introducing a lot of new, rather substantial differences in classes. And there's certain tech things you have to do that can, in the hands of a moderate instructor, will greatly reduce resentment, In the hands of a good instructor, will actually have students much more positive about it.
Number one, they have to not be thinking they're being treated as guinea pigs. OK. So you've got to bring them into the process. You've got to talk to them about why you're teaching this way. Show them a bunch of data. Why you're convinced that they'll learn more effectively from this. You've got to return to that discussion periodically through the term and give them a chance to raise concerns and address those concerns as to what they're learning and why.
So that's pretty much it, really. But they always have to see you're doing this because you're concerned with their learning, and that what you're doing is not just your idiosyncratic, bizarre way of teaching, but really based on other, more fundamental things, and some data that's behind this.
SPEAKER 6: Can I just add one more thing to that? That's great. That's really, very helpful. But I think that the piece that's really important is that the administration-- those student evaluations are really critical, that administration has to be on board.
CARL WIEMAN: That is an important point, because, like I said, you can help a lot, but it's still not at all uncommon for the first year or two of changes for the evaluations to go down. And then recover. And so the administration and so on has to recognize that, as well.
SPEAKER 7: Assuming you're talking about introductory science courses, as you are, have you found any correlation depending on the mathematical rigor of the subject?
CARL WIEMAN: Now, when you say mathematical rigor, do you mean like whether it's physics or biology? Or do you mean whether it's first-year physics or fourth-year physics?
SPEAKER 7: Well, I said introductory courses, so it's the first.
CARL WIEMAN: OK. So I guess probably the best comparison I can give you-- this is on what's the relationship of the mathematical rigor of the subject. The best comparison I can give you is there's lots and lots of data in introductory physics and so on. We recently, at the Colorado Science Education Initiative, went through the transformation process of the third-year E&M course in physics.
Now, I don't know if you're a physicist, but if you aren't, that is the course in which we really think about this is the time at which students really start having to learn some hard core mathematics and applying it to physics, sophisticated mathematical problem-solving and so on. And so, that was designed basically just what I talked about here. But going through carefully and thinking, OK, what within this sophisticated rigorous use of mathematics, what are the things experts do, and we'll make those very explicit to the students and so on.
And the result was actually, it was surprisingly good. It was a dramatic improvement on a bunch of different measures of student mastery of these topics. And really, we were applying the same basic principles that had been proven in introductory physics, up to something where there was a high level of mathematical rigor. And it seemed to work in just the same way. So that's the best particular empirical comparison I can give you.
SPEAKER 8: Just one more question.
CARL WIEMAN: OK. Yeah.
SPEAKER 9: Class structure, at Cornell we have some introductory and upper-level courses that have an [? audio-tutorial ?] format, which you might think actually allows people more of an opportunity to be the expert thinker and get feedback and have a dialogue. I'm wondering if you have any opinion on that.
CARL WIEMAN: Can you explain to me more of what is happening in this course?
SPEAKER 9: Well, there is on in physics. There's an introductory biology and biochemistry. I don't know if people are here for most courses. [INAUDIBLE].
SPEAKER 10: The physics component, it's almost like they have access to hands-on labs the entire time. So they're a series of demonstration stations that they go to, and they can different concepts. And then they have periods that they have several experiments that they're supposed to do. And they go and they do the experiments, and they write it up in their lab book, and then they talk to a TA. And they have specific, assigned TAs.
But they're allowed to do it in their own time. They have a series of homework question that they're supposed to go through, and the solutions are available, too. And they have a testing center, and they're allowed--
CARL WIEMAN: Oh, yeah. Right. This is called the something-or-other. This was big 30 years ago. It got so much work people stopped using it so much. I'd forgotten the label on that. So this kind of self-paced sort of business, it certainly has some features which are attractive to it. I think it clearly has some downsides, beyond just the resources. The resource demand is why a lot of people who started using it don't anymore.
But I think there's no simple answer to whether that's good or bad, because it's really within the details of the implementation, the details of what students are doing. They could be learning to memorize recipes, et cetera. Or they could be learning good expert thinking. And I think the format could easily accommodate either. And so I don't think it's possible to make a general statement about it.
Nobel Laureate Dr. Carl Wieman emphasizes the importance of making science education effective and relevant for a large and diverse population. The approach, he says, is to transform how students understand and use science, and this calls for teaching them to actually think like scientists. Sponsored by Cornell's Center for Teaching Excellence.
