o innovative teaching techniques and investigative laboratories really lead to improvements in student performance? Faculty members who have adopted them think so. But that's hard to prove without objective measures of student outcomes.

The first question to be answered is: What should be measured? Is simple exposure to biology classes and laboratories a useful indicator of success? Should the breadth or depth of student knowledge be evaluated? What is most important: mastery of a body of knowledge, a student's conceptual understanding, or the attitudes a course inculcates?

The University of California-Los Angeles is among the many institutions grappling with these issues. Over the past few years, several science departments at UCLA have teamed up with the psychology department to assess new science teaching approaches more rigorously. "When we evaluate our own programs, we invariably succeed," says chemistry professor Orville Chapman, who, as associate dean for educational innovation, has been leading an effort to incorporate educational technologies into campus science departments and evaluate their effects. "When cognitive scientists evaluate our programs, the initial evaluation is usually rather bad. . . . I urge anyone seriously interested in producing really good learning programs to have independent evaluation by cognitive scientists."

One of the assessments at UCLA compared two sections of a laboratory. The students in the first section were taught with the aid of computers; those in the second section were taught solely by human instructors. The most significant finding, according to Chapman, was the difference in attitudes between the two groups of students. Almost all of the students in the first section were much more enthusiastic about the course and about studying science. Among women, the difference was even more pronounced. In fact, all of the women who took the computer-based course thought they learned more and rated it more highly than other laboratory courses they had taken. "It's the first time I've ever seen UCLA students agree on anything," says Chapman.

Yet when the students taking the two different sections took the same tests for the course, the results showed no difference in learning. "What that says to me," Chapman concludes, "is that we haven't learned how to measure what matters in education." The tests used in the course largely measured learned facts, whereas the computer laboratories focused more on developing skills such as problem solving. "It's very hard to probe understanding in a systematic way that you can document," says Chapman. "So even though the students thought they had gained in understanding, the course did not assess for that."

At Lousiana State University, Marshall Sundberg and his colleagues have had greater success measuring the results of new teaching styles by focusing on concepts rather than factual knowledge. In the process of shifting from traditional to investigative laboratories, Sundberg and his colleagues found that the new laboratories have produced a marked increase in students' understanding of basic biological concepts. Yet in this case student attitudes about the new course were mixed. "It depends on when you hit them with the questionnaire," says Sundberg. "Early in the semester the course is really frustrating for a lot of students. The tools they've developed for studying don't work real well. But in the second half of the semester we see student satisfaction coming back up, and by the end of the course a lot of students love it."

UCLA's Chapman calls assessment "the crabgrass in the academic lawn," and much more work must be done before cognitive research can offer detailed guidance to college faculty. But a reexamination of assessments often has an unexpected dividend, report faculty members who have spent time on the issue. Thinking about assessments forces people to consider the goals of undergraduate biology education, which is why changes in assessment often drive more far-reaching changes in education. "If we don't change the means of assessment, we had better forget about reform," says Chapman.

A new introductory course at Bryn Mawr College demonstrates the subtle links between learning and student attitudes.

• The Laboratory Experience
• Different Approaches
• Making the Transition
• A Survey of Labs
• Undergraduate Researchers
• The Electronic Laboratory
• The Interactive Lab Manual
• The Challenge of Assessment

• A World Of Opportunity
• Changes in the Classroom
• Faculty in the Crossfire
• Women and Minorities
• The Many Paths to Success