CONTACT: Stanford University News Service (650) 723-2558
New education dean enthused about math, science education proposal
STANFORD -- When Richard Shavelson was dean of education at the University of California-Santa Barbara, he came up with a way for science departments to stop their freshman and sophomore majors from switching to the social sciences.
His proposal, similar to one suggested by Stanford's Commission on Undergraduate Education, was to create a science curriculum in which the emphasis would be on teaching students how to conduct a scientific inquiry through which they would acquire specific subject content. The difference is that Stanford's proposal for non- science majors has a chance of succeeding, Shavelson says, because prominent campus scientists are involved in designing it. In the other case, the chief academic officer was a National Academy of Sciences chemist who felt reaching out to ill-prepared students was inappropriate.
While science faculty interest is essential, Shavelson says that there is also expertise to be offered to the campus by the School of Education. Professor Lee Shulman, for example, "is in high demand around the country. He is working with a consortium of universities whose faculty are thinking about what they would like their students to be taught, much as the issue is being raised here."
Shavelson, who will be invested as the eighth dean of Stanford University's School of Education at 4 p.m. Monday, Feb. 13, in Cubberley Auditorium, is a leading expert in the measurement of individual and group performance in education and training. His work in recent years about how ability in science and mathematics is measured suggests that testing methodology influences what is taught and can steer capable students away from learning the skills that are the most valued in academia and the larger world.
In the United States, tests are "the bottom line by which educational quality is judged," Shavelson said, which is why he thinks that new ideas for teaching science at Stanford must be accompanied by new ways of measuring students' achievement in the new courses.
"If one doesn't begin to conceive of how to assess the achievement of students in a way that sends a message that the ability to solve problems in groups over extended periods of time is important - that one doesn't work in a laboratory solely by oneself and have one shot at it - that there is more to doing science than completing multiple-choice tests or solving 20 problems in an hour, then we are sending a message that the new curriculum is really business as usual. That's why in K-12 education, we are involved now in a reform of how we measure achievement."
The Stanford proposal comes at a time, he added, when the quality of higher education is being questioned because of rising costs. Last year a congressional hearing was held about the idea of introducing standardized tests to measure quality, and consumers are grabbing up copies of magazines, such as U.S. News and World Report, that provide their own reports cards on colleges and universities.
"My greatest fear is that somebody will try to measure quality with multiple-choice and short-answer tests, spoiling higher education just as over-measurement in K-12 education has sent the wrong signals to teachers and students."
Fills two-year vacancy
Shavelson comes to Stanford from the University of California-Santa Barbara, where he was dean of the Graduate School of Education since 1987. Previously, he directed the Education and Human Resources Program at the Rand Corp. and was a professor at the University of California-Los Angeles. He received his doctorate in educational psychology from Stanford in 1971.
"He has been a significant contributor to the field of research on teaching for 20 years, helping to move the field into the much more 'cognitive' orientation that continues to characterize it," said Lee Shulman, the Charles Ducommun Professor of Education, who co-chaired the search committee. Shavelson is also versatile, Shulman said. "He's one of those rare people who is equally at home discussing highly esoteric mathematical models of assessment and the nitty-gritty challenges of applying educational policies at a local school district."
Stanford's School of Education has been without a permanent dean since 1993, when Marshall Smith left to become undersecretary of education. Professors Nel Noddings and Myra Strober served stints as acting dean of the school of approximately 40 faculty members and 400 graduate students. The school also draws a number of undergraduates to courses because of their interest in education.
A self-described "entrenched University-of-California guy," Shavelson said he doubts he would have ever left were it not for the state budget crisis that sent three waves of early retirements crashing through his campus. "The last round dug deeply into the people age 55 and older who were ready to be campus leaders, and U.C. lost an investment in its future," he said. Too young to qualify for early retirement himself, Shavelson nevertheless decided it was time to consider returning to his alma mater, Stanford. His wife, Patti, a real estate agent in Santa Barbara, works part time at Stanford's Housing Office. Their two daughters stayed in Santa Barbara, where they are both university students.
"I think of Stanford as the probably the greatest university in the world and quite clearly, the best education school in the world," Shavelson said. He adds quickly, however, that the University of California will be a contender for those honors over the long haul.
"The U.C. system is robust and it will rebound, but it's going to take a commitment to rebuilding, and right now, if you look at the state's budget, the only flexibility is about 14 percent [of the total], which is the universities and prisons. A lot of us believe that without restructuring the state's budget, the University of California will continue to be squeezed for a long time."
Stanford, too, faces budget problems, which is why the new dean of education says he is delighted that President Gerhard Casper's Cabinet is made up of academic deans who advise him on so-called non- academic decisions.
"My experience at the University of California was that the first reaction to budget cuts is to pass the costs from the non-academic departments or areas to the academic ones, where they went down to the faculty, and then the faculty passed them onto the students. It's that kind of administrative decision that, with some experience, deans are going to see coming and guard against."
Math, science education research
By no means, however, does Shavelson plan to spend full time on administrative matters. He brought with him about $450,000 in annual research projects and plans to teach also. "For sanity's sake, I can only administer so much."
His current research involves exploring alternatives to multiple-choice achievement tests in both the schools and the workplace. One of his projects for the National Academy of Sciences involves problems of measurement of military personnel performance while other work involves studies in California's public schools. He also has conducted research on education policy, including the use of information technology in instruction, and has served on a number of advisory commissions and panels at both the state and national level.
Research on educational measurement is both challenging and humbling, he said, because "I've learned over the years that education is an issue of values and politics. The science and scholarship we do can only inform the debate, but it certainly cannot drive it or set the parameters."
In the United States, test scores drive many decisions of government, business and individuals, he said.
"Our culture has become so accustomed to achievement testing that citizens do not even ask how achievement test scores were derived - but only who is on top and who is on bottom," Shavelson wrote in a 1990 article for the journal Phi Delta Kappan. "Achievement scores fit the American belief in a single dimension of ability that distinguishes winners from losers."
Yet Shavelson and others have found the ubiquitous multiple- choice and short-answer tests measure some abilities well and others not at all.
"They emphasize recall of facts and performance of isolated skills but tend not to measure students' conceptual understanding and problem-solving skills," he said. "Consequently, the tests work against what many people value as education, which is application of concepts and procedures to real-world situations."
California's assessment dilemma
In the early 1990s, Shavelson was involved in the effort to reform California's Learning Assessment Program to better measure student performance in statewide testing and to give more timely, instructionally relevant information back to teachers. His work comparing science and mathematics performance assessment methods for the National Science Foundation indicated that it was possible to devise better methods but it would take some time to implement them on a statewide scale. This is partly because testing methods need to be more fully developed with an eye to both cost effectiveness and validity and partly because research indicates student learning can be highly dependent upon the context of time, place and particular investigation to which he or she was exposed. The new methods in science and mathematics assessment involve devising both better questions for paper-and-pencil tests and other methods such as direct observation of students doing science experiments, computer programs that keep track of how students derive answers, and evaluation of students' long-term project portfolios.
"Unfortunately, as our scientific ideas [for assessment] got translated into legislation, the reforms needed to be done yesterday," Shavelson said. "We had envisioned a 10-year project to phase in new methods, but California wound up legislating a performance assessment technology that didn't yet exist. That, coupled with concern in the social sciences that some of the language and literature questions were invasive of the family, dealt the program a death blow. There will be no California Assessment Program at all this year."
Some in Sacramento want to return to the old multiple-choice testing system next year, Shavelson said, but he is involved now in devising a compromise.
"We can't yet provide individual-level information for all California students with new technology. I still would like to see us use it at the school level with a random sample of kids. That would help us send the message that what is important is not just the multiple-choice memory stuff but performance on rich, deep mathematical problems and scientific investigation," Shavelson said.
Sending this message is important because Shavelson also has found that teachers and textbook publishers adjust to what tests measure.
"Eighth grade science books, for example, devote a half page each to every major concept I can conceive of, and part of that page is a picture of a scientist," Shavelson said. "The material basically is introducing a set of vocabulary words so that students can answer any definition that might be included on a multiple-choice answer test. A study by the National Science Foundation found that there is more new vocabulary introduced in science curriculum than in foreign language curriculum."
Yet, he said, the National Academy of Sciences is writing science standards now and "one of the criticisms is that [the draft standards] include everything, even though we can't teach everything. As a country, we can't seem to decide realistically what constitutes science in a science curriculum."
A few years ago Shavelson said he came "face to face with the dysfunctional impact of testing" in his own family. His daughter, who will soon take the Medical College Admissions Test, was then a high school student preparing for the Scholastic Aptitude Test.
"She studied test after test, learning decontextualized vocabulary and tricks to speed up responding on the quantitative sections," he said. "Only when a colleague from Australia remarked that it seemed like a terrible waste of Karin's time to be studying stuff that would be of no use to her beyond the aptitude test did the absurdity of the whole thing dawn on me. She spent untold hours reviewing for the SATs, while ignoring her calculus, physics, English and other academic work."
In his research, with Neil Carey of the Center for Naval Analysis and Noreen Webb of UCLA, Shavelson found that high school students who allegedly understand such textbook equations as F=ma are unable to use the knowledge to plan the solutions to problems. Students often believe that the goal of solving physics problem is to find an appropriate formula into which to stuff numbers, rather than to reason from principles, and they can listen to a presentation on photosynthesis without grasping that plants produce their own food.
While many scientists at the university level are appalled by such findings, Shavelson said that K-12 reforms are "partly being blunted because of the [continued] demand of university professors for students who have learned to calculate quickly and who have memorized all the equations. Some college professors want students who calculate quickly and have a large store of factual knowledge, even though that is not the way the professors themselves do science.”
Shavelson terms this the "gate-keeping" approach to teaching college-level mathematics and science. "There is a tremendous wastage of talent and creative ideas because we tend to admit a few people who look just like us to pass through these gates. We don't get the divergent point of view that we probably need to enrich our scientific endeavor," he said.
What would Shavelson do to make science comprehensible to more students?
For one thing, he said, he would introduce students to laboratories where they would find out that science is not just a matter of getting the right answer.
"When I was in college, I was graded on getting the right answer. Nobody talked about the fact that I should expect a distribution of outcomes from the same experiments done by different scientists or students, and we never debated how it could be that people got different findings but most people got a certain finding. That is, there is a probability of outcomes, but we never discussed this probability distribution and the result you got was [regarded as] either right or wrong."
In contrast, he said, a student laboratory that wanted to produce more scientists would provide "opportunities for students to construct knowledge in a community of scientists and to debate the findings, not just a see-and-show replication of what I told you was true."
This is an archived release.
This release is not available in any other form.
Images mentioned in this release are not available online.
© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300.