Candid Camera

In the fall of 1994, a group of researchers sent a Los Angeles-based videographer in search of American education. Over five months, the videographer drove more than 5,000 miles-from California to Maine to Florida-visiting almost 100 junior high and middle schools along the way. At each stop, he set up his camera and recorded a single 8th grade math class. When the class was done, he dropped the tape into an envelope bearing the address: James Stigler, Department of Psychology, UCLA, Los Angeles, California.

In an ad hoc screening room in the University of California at Los Angeles’ psychology department building, Stigler and a team of international collaborators puzzled over the images on the tapes. Stigler was leading a $2 million U.S. Department of Education project to videotape and study teaching in three countries: the United States, Germany, and Japan. Every tape yielded some surprises, but the biggest shock was the contrast between American and Japanese teaching. The recordings from Japan confirmed the researchers’ suspicions that, because of that country’s national curriculum, its math classrooms are remarkably similar. But Stigler and the others were stumped by what they saw on the American tapes: The diversity of teaching styles that they had expected was nowhere to be found. Everywhere, it seemed, American teachers taught using pretty much the same uninspiring methods. Teachers largely drilled their students on low-level procedures. In not one of the 81 videotaped classes did the students perform a mathematical proof, though proofs were common in Japanese lessons. And while most of the videotaped teachers claimed to be using practices recommended by the National Council of Teachers of Mathematics--practices that encourage students to employ high-level thinking skills--the analysts saw almost no evidence of this.

Since their limited release in 1997, the tapes have become Exhibit A for those who blame teachers for the sorry performance of U.S. students in mathematics. But Stigler and James Hiebert, an education researcher from the University of Delaware at Newark who has worked on the study since its inception, see things differently. American teachers, they say, have little choice but to teach the way they do. They are constrained by a decades-old ethos of schooling so powerful that breaking free of it would require superhuman effort.

In a book to be published this summer called The Teaching Gap, Stigler and Hiebert put it this way: “The widely shared cultural beliefs and expectations that underlie teaching are so fully integrated into teachers’ worldviews that they fail to see them as mutable....It is no wonder that the way we teach has not changed much for many years.”

Stigler and Hiebert believe that the means to improve American teaching are at hand. Collaboration among teachers can steadily improve the classroom lesson, which is the heart of education. Public pressure to overhaul schools gives administrators ample reason to rethink teachers’ schedules and build in the time they need to work together to study classroom strategies.

But for any of this to occur, teachers must discard the notion that the classroom is their private domain. Teaching is too important to be carried out behind closed doors, Stigler and Hiebert contend. For American education to improve, it must become the object of professional scrutiny and comment. And videotaping teachers at work, the two researchers argue, is the first step.

Stigler’s videotape project has long since outgrown his cramped quarters on the UCLA campus. Today he and his collaborators work out of what they call the Lesson Lab--a converted warehouse on Santa Monica Boulevard, just a couple of miles from where old Route 66 runs into the Pacific Ocean. Inside, research assistants peck at the keyboards of a row of multicolored iMacs, working on a new $10 million video study of more than a thousand math and science lessons from seven countries. Computers in an air-conditioned closet hum with more than a terabyte of images and analysis. Education researchers from all over the world, meanwhile, toil in brightly hued cubicles, helping Stigler and Hiebert impose order on the messy, complicated, intensely human act of teaching.

It’s all a bit overwhelming, Stigler admits, given that the project grew out of a casual suggestion made eight years ago. He’d been invited to Washington, D.C., to help design the Third International Mathematics and Science Study, or TIMSS, the largest international survey to date of student achievement in math and science. TIMSS organizers assumed that U.S. students wouldn’t do well on the performance assessments that are part of the study, he recalls, and they wanted a teacher survey to help them understand why. Stigler objected. “I said, ‘I don’t think you can do this with a questionnaire, because we don’t have a shared vocabulary about teaching. What one teacher means by a term like problem-solving is not necessarily what another teacher means, and the problem is worse across cultures.’”

Stigler returned to California and wrote a white paper that proposed to survey teaching practices using videotape. He and other researchers had used videotapes to investigate the dynamics of individual classrooms. Now, he suggested taping on a grand scale to study a representative sample of teachers.

He didn’t hear anything for two years. Then a friend called to tell him that the Department of Education’s National Center for Education Statistics was planning an international comparison of teaching to be included within TIMSS. The center’s request for proposals, it turned out, consisted largely of Stigler’s white paper.

Awarded the contract, Stigler began working with colleagues at the Education Department and around the country to design the study. From the 41 countries participating in the 8th grade TIMSS assessments, the department chose Japan, the United States, and Germany for the videotape survey. The three countries were economic competitors with contrasting approaches to education, and Japanese students had scored well on previous international comparisons.

The researchers randomly selected the classes to be videotaped in each country from a representative sample of those that took the hour-and-a-half performance tests administered as part of TIMSS. Teachers were asked not to do anything special, and Stigler and Hiebert are confident that the tapes show a fairly typical day in the lives of fairly typical teachers.

The resulting images are breathtaking. Watching the U.S. tapes is like being transported back to your own 8th grade mathematics class. Teachers struggle to keep students’ attention. Kids are as monosyllabic as ever. Sometimes, only the nearly ubiquitous presence of overhead projectors marks these classes as different from those of the ‘50s.

The Japanese and German classes also look oddly familiar. The Japanese students are surprisingly unruly: When a teacher asks, “Do you remember what we did last period?” one student responds, “math.” The German students, meanwhile, look like Americans, right down to their casual, slumped posture.

To ensure the objectivity of the research, the bulk of the videotape analysis was quantitative. The tapes were digitized, transcribed, and translated (in the case of the German and Japanese tapes), and each lesson was put on a single compact disk. Teams of researchers then watched the classes over and over, coding almost every action. For example, they divided lessons into class work (where the teacher was talking), seat work (where students were working), and periods of mixed organization. Outside interruptions like public address announcements also were counted.

The results revealed patterns of instruction unique to each country. German classes, for example, usually begin with the teacher leading students through a fairly high-level mathematical procedure. Sometimes, one student solves a related problem on the chalkboard or overhead, while others in the class offer corrections or advice. But much of the class consists of students working at their desks on problems similar to the one presented by the teacher. According to the videotape analysis, the Germans spend about 90 percent of this seat-work time practicing routine procedures.

Japanese lessons follow a fundamentally different plan. Teachers begin by presenting the class with a fairly difficult problem. Students then work alone or in small groups to devise a solution. After 10 or 15 minutes, the teacher calls on students to present their answers to the class. Gradually the teacher and students work together to solve the problem and uncover the related mathematical concepts.

The level of the mathematics presented in the Japanese classes is higher than in the German classes. In fact, the math is so sophisticated that one of the most remarkable things about the Japanese videos quickly stands out: Some of the students clearly aren’t getting it. They seem disoriented. Their solutions are wrong.

Though such confusion would prompt most U.S. teachers to retreat and repeat their instruction, it is a sign of an effective lesson in Japan. Japanese teachers want students to struggle--indeed, force them to--so that they will think about the concepts being presented and learn them on their own. As an example, Stigler and Hiebert cite a strategy for teaching fractions suggested by a Japanese teacher’s manual. Left to their own instincts, students might simply add numerators and denominators, figuring that 1/2 plus 1/4 equals 2/6. Go ahead and let them, the manual urges teachers, and then talk with them about why that doesn’t work.

Such an approach is akin to what leading math educators in the United States have urged for almost a decade. In 1989, the National Council of Teachers of Mathematics released national standards for math instruction, the first of their kind in any subject. Those standards stress the teaching of math concepts--not processes--and advocate that students do real-world problem-solving.

As part of the videotape study, Stigler and Hiebert examined how well American teachers have adopted the NCTM standards in their classrooms. The results were disappointing. Of the 81 U.S. classes videotaped, perhaps two or three show signs of the standards’ influence. The rest stick to a fairly traditional format. They usually begin with the teacher reviewing previous material, often by checking homework. The teacher then demonstrates how to solve the problem for the day. Seat work is assigned, and students work alone or in small groups to solve problems similar to the one described by the teacher. Near the end of the lesson, some of the seat-work problems are checked, and homework is assigned.

The routine nature of U.S. math instruction is painfully clear in the quantitative analysis of the videotapes. In the tapes from Germany and Japan, more than three-fourths of teachers develop concepts over the course of a lesson. Most U.S. teachers simply state concepts. On an international scale, German lessons are at about the 8th grade level, and Japanese lessons are at the 9th. U.S. lessons, meanwhile, are at the 7th grade level. Even the classroom climate in the U.S. tapes seems counterproductive to learning: Thirty-one percent of the American lessons are interrupted at some point by a PA announcement or a visitor coming into the classroom. None of the 50 Japanese lessons is interrupted.

Most damning of all are the results of an analysis of the taped lessons by four experienced high school and college teachers known within the project as the Math Group. Working from summaries of the classes that hid each tape’s country of origin, the group members judged the lessons in terms of such measures as coherence and complexity. It was their blind analysis of the tapes that revealed that no U.S. class had done a single mathematical proof. “I can’t begin to tell you how appalled I was when I saw that result,” says Alfred Manaster, head of the Math Group and professor of mathematics at the University of California at San Diego.

These findings don’t surprise the leaders of the mathematics reform effort. Though the videotapes were made five years after the NCTM standards were released, math instruction was just beginning to change, says Glenda Lappan of Michigan State University, president of the NCTM. “Materials that can help teachers reach for the standards are only now becoming available,” she says.

But the videotapes don’t lie, she admits. “At the policymaking level, states and districts have been very influenced by the standards. But there aren’t many places where teachers have been supported to move forcefully toward the standards....Many teachers just don’t realize how much of a U.S. model is built into their minds.”

That model does not seem to be doing much for students’ understanding of math. On the TIMSS achievement tests, Japanese 8th graders did extremely well, finishing among the top handful of nations. U.S. students, on the other hand, were solidly in the middle of the pack. Says Manaster: “There is no doubt in my mind or in the minds of my colleagues that the depth of exposure to mathematical ideas in the United States is much weaker than in Japan and probably weaker than in Germany.”

As Stigler and Hiebert neared the end of the videotape analysis, they realized they had a problem. To encourage teachers and students to participate in the study, they had promised that researchers would be the only ones to see the tapes. Because of these confidentiality agreements, their results existed as mere numbers on paper.

With the encouragement of the Department of Education, the researchers began looking for ways to take the videotapes public. In some cases, they were able to secure permission from the teachers and students. They also sent videographers back into the field to retape representative classrooms.

As they began releasing the preliminary results of the study three years ago, they showed this raw footage to illustrate its conclusions. Nanette Seago first saw the tapes at an NCTM seminar. Seago, who at the time was working for a California group that conducts workshops for middle school mathematics teachers, immediately saw the tapes as a valuable professional-development tool.

“I use the analogy of sports,” she says. “Now we can do instant replay in the classroom. We can pause, replay, and analyze an interaction that’s complex and fast and in real time.”

Seago launched a one-woman crusade to make the tapes more user-friendly. In their original form, the tapes simply showed a class from beginning to end. They needed context and editing to direct attention to important aspects. Seago also discovered that the way the tapes were presented had a big effect on teachers. She learned, for example, that if she showed the Japanese tapes first, “teachers may think that you’re showing them the right way to do things, and they may dismiss what they’re seeing by saying that nothing like this could ever happen to them. Then you’ve lost an opportunity to analyze the teaching in the video and use it to help them improve their own teaching.”

Eventually, the Department of Education hired Seago to adapt excerpts from the tapes. The resulting one-hour video, Eighth-Grade Mathematics Lessons: United States, Japan, and Germany, has become a centerpiece of teacher training around the country. Professional-development groups have shown it to thousands of teachers at small workshops and national meetings.

Teachers’ reactions to the tapes have been intense. “The videos can be taken as an insult,” says John Ribeiro, who provides professional development to teachers in the West Warwick, Rhode Island, school district. “It can seem to be a put-down of the American way of teaching, even though that’s not the intent.” To counter such thinking during workshops, Ribeiro always points out a key feature of the American lessons: The teachers are working extremely hard. If they were using different methods--and if they could get their students to work equally hard--their classes would soar, he says.

Some teachers who see the tapes admire their foreign counterparts’ work. “I was very impressed with the focus, the oneness, of the lessons in the Japanese classroom,” says Maurice Page, a mathematics teacher at Cambridge Rindge and Latin School in Massachusetts. Though Page, like other teachers, acknowledges the difficulty of transplanting instructional practices from other countries to the United States, he believes there are important lessons to draw from abroad. “Classrooms in America are dominated by the idea of coverage....The Japanese classrooms are about depth more than breadth.”

A number of educators respond emotionally to the tapes. “I’ve seen teachers with tears in their eyes after watching some of the lessons from other countries,” says one researcher. “It’s amazing to see a lesson play out so well.”

The contrast between U.S. and foreign lessons sparks many of these reactions, educators and researchers say. U.S. teachers hold powerful mental images of what classrooms should look like and how they should be run. Stigler and Hiebert refer to these mental images as a “script.” And they contend that everyone in the U.S.--students, parents, and teachers alike--shares more or less the same script. Classes run according to these shared preconceptions, and few people demand anything dramatically different than what has been done in the past.

In The Teaching Gap, Stigler and Hiebert describe this commonly held mindset about education in sociological terms. Teaching, they say, is a culture, a set of learned beliefs and practices. Teachers soak up part of this culture during their preservice training, but they absorb most of it during their 17 years on the other side of the teacher’s desk as students. As a result, the culture is entrenched and resistant to change.

To illustrate this, Stigler and Hiebert point to American math teachers’ favorite tool: the overhead projector. Math teachers in the United States have used the overhead for years as a way of focusing students’ attention on a single element or concept. Covering the transparency with paper, they can reveal information piece by piece, never showing more than one or two lines at any given time.

Japanese teachers, by contrast, almost never use overheads. Instead, they gradually fill the chalkboard with a running record of the mathematical concepts that they have covered. Unlike the American teachers, they emphasize the connections between concepts and record those connections on the board so students can reflect on them throughout the lesson.

If U.S. teachers were forbidden to use overheads tomorrow, write Stigler and Hiebert in The Teaching Gap, “teaching would not change much. The chalkboard would simply be used to fill the visual aid slot in their system and therefore would be used as an overhead projector-to catch and hold students’ attention.” Most likely, teachers would write one concept on the board, discuss it, and then erase it before moving on.

This tenacity of the teaching culture explains why education reform has faltered, the researchers say. Whatever the popularity of smaller classes, voucher programs, or intensive teacher workshops, they are unlikely to improve learning because they don’t change what happens in the classroom. In other words, they don’t change the culture. In fact, most quick fixes just breed cynicism about reform. “That’s my biggest fear,” says Sega, “that we’re going to have a six-step Japanese plan that would go on posters in every classroom. That’s so us.”

Rather, Stigler and Hiebert argue for a new kind of professional development that would go beyond the typical two-day seminar. “If you have a factory and half of your widgets are coming out defective,” says Stigler, “you wouldn’t send all of your workers off to a weekend workshop on how to build widgets. You’d go through your factory and see what’s wrong.”

For schools, fixing the factory means looking at what actually happens in the classroom, Stigler adds. And that’s where videotape comes in. It can slow and stop the action. It reveals exactly what teachers say and how students respond. If put in digital form, it can give observers instant access to any part of a lesson, allowing searches by keyword or concept.

In The Teaching Gap, Stigler and Hiebert advocate using video for what they call lesson study. Groups of teachers would regularly study individual lessons together-such as teaching mixed fractions to 4th graders, the Revolutionary War to 7th graders, or mitosis to high scholars. They would do research to understand the cognitive abilities of their students. They then would analyze a lesson and determine how to improve it. Finally, they would disseminate the results of their work as widely as possible.

In Japan, write-ups of lesson studies are available in bookstores. Stigler and Hiebert foresee something more high-tech: a library of digitized lessons and commentary, fully accessible to teachers, administrators, and parents via the Internet. Teachers who are beginning a study of a particular lesson could watch dozens of similar lessons from different countries. In turn, they could make their own videos and spotlight their own strategies. In effect, Stigler and Hiebert argue, teachers would become researchers of classroom practice, giving them a professional status that all the certificates and tests and degrees in the world cannot.

Researchers and teacher trainers have already taken tentative steps in this direction. Using software developed by a company that Stigler has spun off from the Lesson Lab, Seago is developing a series of CD-ROMs that combine videotaped lessons with transcripts, student work, and commentary. The videos are of “real teachers trying to improve their practice,” she says.

At the University of Michigan, researcher Deborah Ball has taken a somewhat different approach. For an entire year, she had the 3rd grade math class that she taught filmed every day. She now uses that library of tapes in teacher workshops to explore common classroom problems. “A lot of professional development is very remote from teaching,” she says, “or it’s so concrete that it has no shelf life--it’s all about what you do next Monday.”

The Education Department is eager to use videotape in future studies of teaching. “It’s a breakthrough methodology,” says Pascal Forgione, commissioner of the department’s National Center for Education Statistics. “It allows us to get beyond the horse race [of international comparisons] to the essential aspects of teaching.”

And video is becoming key to teaching’s new national certification process. Candidates seeking a certificate from the National Board for Professional Teaching Standards must submit a video portfolio of their teaching-videos that some would like to see compiled as a library of best practices. “The use of videotape has enormous potential to break up the long-standing cultural imperative of this profession, which is a culture of isolation,” says Barnett Berry, director of the Southeast office of the National Commission on Teaching and America’s Future.

Of course, the biggest hurdle to such grand schemes may be teachers themselves. Most are accustomed to working behind closed doors and would view cameras in classrooms as an Orwellian intrusion. “Being videotaped is not a comfortable human feeling for any of us,” says Ron Gallimore, an educational psychologist at UCLA and a close collaborator with Stigler on the new video study. “Most teachers have never been videotaped, and they’re extremely spooked by the idea.”

They also may get spooked by how the videotapes could be used, says Alice Gill of the American Federation of Teachers, which helped fund the writing of The Teaching Gap. “If teachers think that you’re going to videotape and use them as a bad example, they’re going to resist. Part of that comes from the way evaluation has historically been done in this country-more as criticism than as constructive help.”

For videotaping to become accepted, Stigler and Hiebert argue, teachers will have to see it as a tool, not a threat. Cameras will have to be used to study and evaluate teaching, not teachers, says Stigler. “What you do in the classroom is not only a function of who you are but of what methods you’re using. We have to shift the focus from the teacher to the methods being used.”

Ultimately, videotaping’s potential to improve schools may dispel teachers’ fears. Stigler and Hiebert believe taping classes will help focus school reform on the one thing that really matters in education-the interaction between teachers and students. That alone could convince lots of teachers to throw open their classroom doors and let the cameras roll.