Los Angeles
When James W. Stigler first thought about examining math teaching by studying videotapes of classes in the United States, Germany, and Japan, “it really was impossible,” the UCLA professor recalled.
Anyone with a videocassette player knows how difficult it would be to try to analyze key events or draw broad comparisons while fast-forwarding and rewinding through hundreds of hours of videotapes.
Nevertheless, Mr. Stigler got a contract from the U.S. Department of Education’s National Center for Education Statistics for just such a study. “After I got the money,” he recalled matter-of-factly, “I found out it was really impossible.”
The agreement called for the videotaping of representative national samples of 8th grade classrooms as part of the Third International Mathematics and Science Study.
A difficult task, but not really impossible, said Lois Peak, the Education Department’s TIMSS project director. She said with a chuckle that Mr. Stigler’s comment merely reflected his humility.
Yet surmounting the technological problems involved in the three-year, $2 million study was by no means guaranteed, Mr. Stigler said in an interview here in his office at the University of California, Los Angeles. When he began the project, the technology did not exist to match his vision for managing, on a computer, the large amounts of data the study would collect.
Three and a half years later, that vision is contained on a digital computer database--millions of videotaped images of 231 separate lessons from math classrooms throughout the three countries.
Instant Access
The software, designed specifically for the project, allows a viewer to access any given moment of any lesson and view it on a computer monitor alongside a written transcript keyed precisely to that section of the tape.
The system differs from off-the-shelf CD-roms, which carry someone else’s ideal configuration of images and text, because the user can manipulate the video and text to create a customized document.
Using this technology, Mr. Stigler’s researchers can quickly access video images of American teachers giving a geometry lesson, then follow it immediately with a similar lesson from Japan.
This has enabled them to learn, for example, that 8th grade teachers in Japan give a coherent lesson tightly focused on one mathematical concept, while their U.S. counterparts tend to introduce many topics at a time.
The technology, Mr. Stigler said, has applications far beyond the world of education research. A lawyer could search, say, a videotaped deposition for any use of the word “car,” then annotate that scene with questions or comments.
This breakthrough offers many advantages. ones and zeros read by a computer, the files won’t wear out with repeated playing. And with instantaneous access to any point on any of the videos, the researchers and others can perform much more sophisticated analyses than would have been possible with conventional videotape.
“We really couldn’t have done this study without this technology,” Mr. Stigler said.
The birth and early life of the software proved arduous. As of late 1993, as the researchers prepared to collect all their data, they had experimented with hardware and software combinations without real success. One option, they knew, would have been to wait and let the technology catch up with them.
But Mr. Stigler would have none of that. “I said, ‘We’ve got to do it now. We can’t take no for an answer.’”
So they hung on to “the bleeding edge of technology,” Mr. Stigler said, trying to work out the bugs in the system.
Money for the software project also became an issue. So with the blessing of the Education Department, Mr. Stigler and others founded a private company, Digital LAVA Inc., in 1995 to raise money to continue the software project.
Many Applications
So far, the company has sunk about $2 million into the project, compared with the $100,000 the department put in, Mr. Stigler said. The company, located in a small office suite just a few blocks south of the UCLA campus, has attracted other users of video who want to buy the software, but Digital lava has yet to turn a profit.
The main database software, known as vPrism, runs on Apple Macintosh systems and is priced at $4,498 per copy for educational institutions and nonprofit organizations. As of next month, a desktop spinoff will be available that will sell for $198 for nonprofits and will run on personal computers with Windows operating systems, said Tom Stigler, the company’s vice president for sales and marketing and the brother of James Stigler.
The expense may be a stumbling block for some education researchers, cautioned Harold W. Stevenson, a University of Michigan psychologist and a colleague of Mr. Stigler’s.
But he and other experts acknowledge that the technology has potential.
“Methodologically, this is going to be a whole new direction in our ability to understand and quantify phenomena,” Ms. Peak of the NCES said.
Mary Lindquist, a professor of mathematics education at Columbus State University in Columbus, Ga., agreed. The digital database, she said, “gives us real power to look at classroom instruction in many different ways with many different eyes over a long period of time.”
--Millicent Lawton
