Brains are like muscles -- if you want to beef them up, you have to exert them strenuously for a long time.

Learning, then, takes a great deal of time, practice and sustained effort, said physicist and Nobel Laureate Carl Wieman in a public lecture, Sept. 22, in Cornell's Plant Science Building. Science instructors, he said, need to use the data on how memory works and the same process that scientists use to glean new information: conceptual problem solving.

Wieman, who won the Nobel Prize for physics in 2001, directs the Carl Wieman Science Education Initiative at the University of British Columbia and the Colorado Science Education Initiative. In his talk, "Science Education in the 21st Century: Using the Tools of Science to Teach Science," he urged for reforming traditional science education.

The standard model for teaching introductory science -- with lectures given to a hall full of silent students, homework assigned from the back of a textbook and exams -- is failing, said Wieman. This model enforces "novice" thinking, which Wieman defines as gathering isolated pieces of information, matching patterns and applying memorized recipes. Novice thinking is the "plug and chug" approach to problem solving: finding the right equation and plugging in the right numbers without understanding why the formula works.

"An effective science education transforms how students think," Wieman said. More important than giving students facts and equations to memorize, he said, is teaching them what he calls "expert-like thinking" -- that is, concept-based strategies for problem solving, a mental framework for retrieving and using facts and the ability to monitor one's own thinking and learning.

Acquiring new ways of thinking, said Wieman, takes hundreds to thousands of hours of intense practice with guidance and reflection. Most students are not getting the practice they need.

Carl Wieman follow-up discussion luncheons

To explore what Cornell's teaching staff can learn from Wieman's ideas, two luncheons are scheduled:

• For engineering, science, math faculty

Tuesday, Sept. 30, 11:45 a.m. -1:30 p.m. (come for any amount of time), Duffield 340, lunch provided.

Where do we need more innovation? What do we need (remodeling, resources, rewards) for such changes?

The luncheon is co-sponsored by the College of Engineering's Teaching Excellence Institute and Cornell's Center for Teaching Excellence. RSVP to Kathy Dimiduk, director, Engineering Teaching Excellence Institute, at klc78@cornell.edu.

• For graduate teaching assistants

Thursday, Oct. 2, noon-1:30 p.m., 231 Phillips Hall 231, lunch provided.

This interactive session will address the teaching principles, methods and ideas presented during Wieman's visit. RSVP to cornellcte@cornell.edu. The luncheon is co-sponsored by the Engineering Learning Initiatives, the College of Engineering's Teaching Excellence Institute and Cornell's Center for Teaching Excellence.

Wieman cited studies showing that most physics and chemistry courses leave students with an even more novice-like stance towards science than they had initially. Other studies show that, based on quizzes given at the start and end of introductory physics courses, students on average learn less than 30 percent of the new concepts presented.

Traditional science teaching is ineffective because instructors have been taking the wrong approach, he said. "The experts who teach science courses find the conceptual underpinnings obvious, and they don't understand the limitations on a student's working memory," he said.

Science instructors need to approach the teaching of science like science itself, Wieman said. Teaching practices should be based on good data and standards of evidence, not instinct or tradition. Instructors, therefore, should familiarize themselves with what cognitive scientists have learned about learning and memory, he said. For example, the part of memory that deals with processing information is extremely limited compared with long-term memory.

"A student can only retain four to seven new items before they hit the 'screen of death,'" Wieman said. "There's not enough time in lecture to develop long-term memory. Instructors need to design homeworks to give students many hours of expert-like thinking."

They also need to take advantage of current technologies to engage students in the classroom. "There's nothing genetic about an interest in physics -- if you want them to learn, you need to motivate them," Wieman said. He recommended using clickers -- remote polling devices -- to engage students in lecture and initiate discussions.

Wieman's lecture was sponsored by Cornell's Center for Teaching Excellence.

Graduate student Melissa Rice is a writer intern at the Cornell Chronicle.