There’s a story Biology Professor Ken Yasukawa tells in his Biometrics class about an army in ancient times that needed to ford a river. The army sent out scouts who came back and reported that, “over there, the mean depth is one meter, so we should be able to cross.” In crossing the river, the entire army drowned. Why?
“They didn’t talk about the variance,” Yasukawa tells his class. “So the mean depth is one meter, but the deepest part is 20 meters. So when they hit that 20-meter point, they all drowned. It’s not sufficient to know the average, the mean.”
The story points to the need for proper research design, one of the concepts at the heart of his Biometrics class. The course recognizes that biologists working in a world of sophisticated statistical software and vast amounts of information need to know how to acquire and analyze that data, design experiments, study and make inferences from the data, and report the results.
Yasukawa, holder of the Mead Family Chair in the Sciences, came to Beloit some 30 years ago when the biology department didn’t have a requirement for its students to take a statistical course, though there was an optional offering. Very few students took it.
“We decided as a department that the only way to generate enough interest was to require it,” Yasukawa says. The requirement remains “somewhat unusual” among biology departments, though the way it’s taught is very different, he explains.
In his own graduate days, Yasukawa remembers his biometrics course as “completely ineffective”—straight lecture format, separate lab, biological examples divorced from context or reference.
His students’ post-class evaluations of his first go-round with Biometrics at Beloit—he taught it in the same style as he’d been taught in graduate school—said the same thing.
“They said, ‘We can’t retain any of this, there’s no framework, there’s no context, it’s not computing for us.” Yasukawa took their suggestions to heart. “I thought long and hard about what kinds of changes I needed to make so that teaching Biometrics is more effective.”
What Yasukawa wanted, then and now, is to have his students learn analytical skills that they could apply in biology, to other classes and, more broadly, to the world beyond Beloit. So he revamped Biometrics at the college.
Instead of using a course setup that separated labs from lectures, Yasukawa (and Professor of Biology Yaffa Grossman, who also teaches sections of the class) began employing a workshop style of teaching: two hours, three times a week, always in the lab. Each workshop session starts with a mini lecture, then a hands-on activity informed by the lecture, filling the gap in the previous iteration of the class, which lacked an experiential element.
“It’s hands-on, minds-on,” he says of the current version. “They’re actually engaged in doing something that is informed by the lecture.”
Though he is still using the same erudite, dense textbook, he tells students not to read it from start to finish, but rather to use it as a resource, annotating it along the way and saving it for future reference in grad school. He writes worksheets to direct their attention to key points.
This workshop approach to teaching quantitative analysis has proved effective.
Erika Zevin’10, an integrative and medical biology major at Beloit, now in her third year of medical school at the Medical College of Wisconsin in Milwaukee, sailed through her graduate-level statistics course, Evidence-Based Medicine. She credits Biometrics at Beloit.
“I can definitely say that my experience with Biometrics at Beloit helped me ace that class,” she says. She achieved a score of a 100 percent in the graduate course. “As for non-scientific areas of my life, Biometrics was a class [in which] I felt I had to take responsibility for my learning; I had to figure out the answers to my own questions. This skill of individual responsibility is valuable for every aspect of your adult life.”
Felicia Hernandez’15, a health and society major, took Biometrics during the fall 2012 semester and appreciates that her coursework not only taught her to analyze data, but to peel back the layers of forming, conducting, and measuring that go into a research study.
“Before enrolling in this course, I did not understand the complexities in gathering evidence and evaluating research. Analyzing data is a specific process that is complex and detailed,” she says.
Though the class is composed of 70 to 80 percent biology or biochemistry majors, it also draws students from psychology, geology, and health and society. The class examines studies from the field of biology, but Yasukawa also directs students to non-discipline-specific examples, so routine they can be found in a daily newspaper—he points to an article about school district test scores as an example, citing its misuse of statistics. As a class assignment, he sends students to the library to find journal articles with improperly used statistics, an assignment that usually winds up being much easier than students anticipate.
“I think anybody with a liberal arts background should be able to argue a point using quantitative evidence—should be able to read a graph, be able to make a graph, and present it persuasively,” Yasukawa says.
Biometrics in a nutshell: what students do
- Search and read the literature
- Identify and formulate questions and hypotheses of interest
- Propose research
- Implement research
- Report results
- Evaluate the work of others