A standing room only crowd filled the University Club’s largest meeting space to hear Nobel laureate Carl Wieman, a noted expert on science education research, share his wisdom on science teaching and learning.
Wieman, a professor of physics and education at Stanford and former associate director of the White House Office of Science and Technology Policy, spoke on “Taking a Scientific Approach to Science Education.”
The Jan. 30 talk was the inaugural event of Pitt’s new Discipline Based Science Education Research Center (dB-SERC), led by physics and astronomy faculty member and physics education researcher Chandralekha Singh. (See Jan. 23 University Times.)
In welcoming remarks, Dietrich School of Arts and Sciences Dean N. John Cooper said the launch of dB-SERC springs from the University’s “renewed commitment to the centrality of the undergraduate program in so much of what we do.”
Cooper said, “The ways in which one can successfully teach undergraduates the principles and practices of the core disciplines are changing all the time, and changing in a way that can be informed by evidence-based experiments on what works better in the classroom. Within the school we have been talking about ways in which we can make sure our programs are cutting edge, that they take full advantage and enhance the qualities of a major research university and that they truly prepare students for careers in the 21st century.”
The center’s purpose “is to help faculty in the science departments to explore and figure out better ways of teaching their disciplines at all sorts of different levels and to translate that into the undergraduate programs,” Cooper said.
“I think it will be an enormous help to individual faculty but it’s also going to be a very valuable resource for chairs and departments who are working with the various deans to really transform the quality and impact of the undergraduate program.”
*
A key goal of science education, Wieman said, is “to have, in all the different fields, students better understand and appreciate and be able to think about the respective areas of science, engineering and math … more like the scientists in that field” — even if they don’t go into that field.
But what is it that instructors in the sciences want their students to learn? What does it mean to help students think more like scientists? And how is that achieved?
New research is changing commonly held ideas about how people learn complex tasks, said Wieman, who joined the Stanford faculty in September.
The old paradigm — that the extent to which knowledge is absorbed into a student’s brain is related to differences in individuals’ brains themselves — is being replaced by the view that students’ brains actually aren’t all that different from one another.
“The learning process is really them transforming, developing via suitable exercise,” Wieman said. “And to the extent that they undergo similar kinds of exercise, they achieve similar kinds of learning.”
Wieman said that when he first started teaching physics more than 35 years ago, his approach was similar to most: Think hard about the subject, figure it out clearly, then explain it to students so they understand it the way the instructor does.
“I’d give them some problems to solve and if they could do that problem — great; they’d mastered it and we were done. If they couldn’t do the problem, it must mean there was something wrong with the students because I clearly understood it and I’d just explained it to them,” he said.
Unlike “mean” faculty, whose solution was to simply get rid of the bad students, “I tried telling them louder with the expectation that it would achieve better results,” he quipped, drawing laughter from the audience.
The situation was frustrating. “It was clear to me that my brilliantly clear explanations were leaving the vast majority of the students quite baffled.” His colleagues weren’t doing much better, he said. “It seemed like it was just kind of a reality I couldn’t do much about.”
Wieman said his enlightenment came over the course of working with graduate students in his atomic physics lab.
“These students didn’t get into my lab unless they’d had 17 years of being very successful in the classroom, particularly in physics courses. But in my research lab, they were pretty much clueless about how to actually do physics,” he said.
“But after just a couple of years of working in the research lab, they had turned into expert physicists.”
Over time, he found a consistent pattern. “If anything, there was this anti-correlation between success in coursework and how good a physicist they became. I became convinced that there was really some fundamental question here.”
Wieman set about tackling the question: “What do we know about the research on how people learn, particularly about how they learn physics?” He studied existing research then did studies himself.
In the past couple of decades, major advances have been made in cognitive psychology (studying how people think and learn); brain research, and the study of college science classrooms, Wieman said.
“The findings from all three of these areas are now coming together in a very nice way to give us a much clearer understanding of guiding principles about what’s important and how to achieve learning of complex expertise like science and mathematics.”
Expert thinking
Cognitive psychology research has revealed three basic elements in what makes up expert thinking, he said:
• Experts have lots of factual knowledge about their discipline.
• Each discipline has an organizational mental framework used to retrieve and apply that knowledge to solve problems.
• Experts have the ability to monitor their thinking and learning, checking themselves on their understanding and on ways to arrive at solutions.
Research has shown that no one innately has these characteristics in any discipline, Wieman said. “These are fundamentally new ways of thinking,” he said.
Expertise is built much in the same way that exercise builds up a muscle. “Everybody requires many hours and, to reach a high level of expertise, thousands of hours of intense practice actually to develop expert-like abilities,” he said.
“The brain is substantially changed and rewired in this process to develop expertise,” he said. “Biology says you’ve got to put in this much work to develop it.”
Putting in the hours is important, but how those hours are spent matters too. The learner must be undertaking challenging but doable tasks that provide practice at expert thinking.
Learners also must get feedback on their performance and reflect on it, he said.
It takes about 10,000 hours of this intense practice to develop expertise, after which the learner has a very different brain, Wieman said.
That makes it hard for experts to understand and perceive things the way novices do, he noted.
“One of the challenges of teaching is that you’ve got no self-reference for your brain. So you think that is how you learned. It’s not actually correct at all because you’re taking a different brain and trying to project it back.”
Effective teaching
What does a good teacher do? “The best we understand is to really think of the teacher as a cognitive coach,” he said. “This is what the learner has to do; a good teacher facilitates this.”
Teachers must identify the sort of thinking that makes up expertise and design practice tasks that will develop it. They also must provide useful, specific feedback that guides students on how to improve. And they need to motivate their students.
“You really can’t expect a science student in a subject they know nothing about, to see ‘Oh, it’s really important that I put in a thousand hours working intensely on this when I don’t even know what it is.’ The teacher has to also show the learner as to why this is important and useful and rewarding for them to put in all that effort.”
Effective teachers must have high levels of expertise in their subject, he said. “At some level, this is really the justification for research universities. It’s why expertise is important in being effective at teaching.”
Evidence from the classroom
How to apply the principles varies by subject area, Wieman said, but effectiveness of alternative teaching methods generally is measured against the standard lecture approach.
One area that’s been studied is conceptual mastery: How well students take the underlying concepts or models used in the discipline and, given a new situation, correctly recognize which concepts apply and use those concepts to make correct predictions on what will happen in that situation.
In physics, for example, a common measure is the force concept inventory, which takes a subset of the essential concepts of force and motion and measures how well students apply them to real-world applications. Given to students before and after a course, the inventory provides a measure of learning produced by the course.
Wieman said research has shown that in traditional lecture courses the amount students learn typically doesn’t exceed 30 percent of the essential concepts teachers want students to learn. “They only learn a third of what they didn’t know coming in,” he said.
The findings are independent of other factors such as the quality of lectures, class size or institution. “The traditional lecture approach is simply an ineffective way for a student to develop conceptual mastery,” he said.
Teaching that includes ways for students to actively practice concepts shows results higher by a factor of two, compared with traditional lecturing, Wieman said. “This is a result that’s been replicated over and over and over.”
He cited several examples:
One study of introductory physics at Cal Poly showed that switching to a “studio physics” approach in which the instructor facilitated and coached students in carrying out a series of learning activities yielded better conceptual mastery.
“The student learning was really dominated by the teaching practice the instructor used, not any other characteristic,” he said, noting that some instructors who simply applied this different method of teaching found their students learned six times more.
“That’s pretty incredible,” he said. “How can any university administrator look at that and not go back and say to faculty: ‘You’ve got to teach this way. We can’t tolerate our students learning a sixth of what they might’?”
A University of California-San Diego study of computer science learning showed that a peer instruction approach (in which students talked among themselves and responded with clickers to questions posed in class) cut failure and dropout rates from 20 percent to 7 percent.
Wieman noted: “This is a whole lot of students who now are successful in the subject going on … that would have failed or dropped out otherwise.”
Noting that such studies measure the learning that takes place over the course of a term — including what might be gained by doing homework and studying for exams — another study pitted an experienced, successful professor against a postdoc trained in the principles of effective teaching to better study the impact of in-classroom learning in different 250-student sections of the same physics course.
The two instructors agreed on the specific learning objectives, covered the same material, jointly prepared an exam and had the same amount of class time.
The result: Students taught in the traditional section scored an average of 41 percent on the exam compared to an average of 74 percent in the classroom where principles of effective teaching were put into practice.
The results answer the question of which students these teaching methods benefit: Are they just for the best students? Or do they favor struggling students at the expense of better ones?
Displaying a histogram of scores, Wieman said, “That entire distribution moved up: It says this is a really much more effective teaching method for everybody with a human brain. That’s really what you’d expect, because these are principles really about how the brain works, not how the well-prepared brains learn versus poorly prepared brains learn.”
Simple practical applications
Wieman concluded his remarks by sharing three basic findings that professors could apply in class immediately.
• Integrate expertise-building activity into homework and exams.
Although most learning in science courses comes through doing homework and studying for exams, “the typical homework problem is not teaching or testing expertise at all,” he said.
Exams and homework problems usually give students only the information needed to solve the problem and tell them which factors to ignore, eliminating the opportunity for students to practice separating relevant from irrelevant concepts.
Students typically are asked only to provide an answer, with no argument for why that answer is reasonable, wiping out that opportunity for developing expertise in evaluation.
And, although each area of science has its own set of representations among which experts move fluently, students typically are asked to provide only one representation: “Produce this graph” for example, eliminating the opportunity for them to practice moving from one representation to another.
In most cases, students often find it more efficient to simply memorize the “right” procedure that they know they’re being tested on “without having to decide whether it’s appropriate or not, because we’re structuring the course so they don’t have to,” Wieman said.
“Think about the nature of expertise in your subject and make sure it fully is represented in the practice tasks that your students are going to do.”
• Recognize the limitations of short-term memory.
Memory in humans isn’t only about remembering over long periods of time, Wieman said. Short-term working memory “remembers new things and processes them over short time scales, like would be important in a class, seeing material for the first time,” he said.
Unlike long-term memory, working memory has a very limited capacity, he said. “And it doesn’t just remember things. It’s a CPU. It processes things as well. And the more it’s called upon to do, the more it slows down and can’t process and learn as well.”
To protect students from the equivalent of the Windows “blue screen of death,” instructors must recognize that short-term memory can handle five to seven distinct new items, Wieman said.
“By ‘new,’ that means things that aren’t already in long-term memory,” he said, noting that is just a tiny fraction of what’s presented in a normal class period. “What this means is that you should really think about anything that puts unnecessary demands on working memory.”
Reducing jargon and nonessential information are easy adjustments instructors can make. “Those interesting little digressions … have real cognitive costs. They really reduce learning so you’ve got to recognize those costs. Limit your information so it’s not overwhelming.”
When an instructor tosses out a technical term, “You really should think to yourself: Is that one-sixth of what I want students to get out of class tomorrow?”
Instructors also can help students by explicitly showing how the information is organized and connected together. If students can make the connection, two concepts can be fused together, taking up less working memory, he said.
Tying new concepts into things that already are in students’ long-term memory — by using analogies, for instance — cuts demands on memory and helps students learn better, he said.
Instructors should plan to give students’ working memory a break during class, he said. “The very simple operation of stopping a lecture with posing some challenging conceptual question to the class where they have to talk to each other and indicate an answer” is beneficial, he said. “And it doesn’t actually matter whether they can solve the problem or not, just that they get engaged in it.”
Not only can students help one another clarify concepts they may have missed or forgotten, but they’re practicing expert thinking by thinking hard about a question, he said.
“When someone is engaged in a question, even if they can’t answer it, it’s tremendously beneficial for learning from a follow-up explanation,” he said. “It’s tremendously more powerful than giving them the explanation before they know the question.”
• Be attuned to what students understand.
Instructors should listen in on students as they work together in class. “Just listening in on those discussions, you get so much better at finding out where they’re at, what they’re struggling with, what they understand. And you’re poised to be much more effective in your subsequent follow-up discussions and presentations,” he said.
In closing, Wieman recapped his vision: “Approaching teaching and learning really as a science, doing research, looking at these basic principles of how people learn, thinking about applying these in the classroom and testing how they work. And so we can really make this like astronomy rather than like the astrology that, frankly, much of our approaches to teaching have been in the past.”
___
Kimberly K. Barlow is a writer for the University Times at the University of Pittsburgh. This article is reprinted with permission from the Feb. 6 edition.
Subscribe to our monthly newsletter.