Outstanding Teachers: Two Professors’ Work Brings Statistics Into the Light

April 28, 2013 | Salute to Scholars, The University

Great Teaching is at the heart of a great university. Nine University professors featured here have received special acclaim for their instructional acumen from Carnegie Foundation, Presidential and Chancellor awards to recognition by The Princeton Review. Their classroom magic inspires students and prepares them for the future.

covMATHPROFSam01

BILL WILLIAMS AND SANDRA CLARKSON • MATHEMATICS AND STATISTICS PROFESSORS • HUNTER COLLEGE

BILL WILLIAMS jokes that what led him to collaborate with Sandra Clarkson was the constant refrain at cocktail parties: “Oh, you teach statistics? I hated statistics!”

The former Bell Labs research statistician understands why. “You’d be floored if you looked at a first-year statistics textbook and got out the very first one written by George Snedecor in the 1930s. They looked the same, the chapter titles were the same and the teaching of statistics hadn’t changed. Sandi and I set out to do something different.”

The software-based, graphics-heavy solution devised by these two Hunter College professors since the 1990s has made “Elementary Probability and Statistics” amazingly popular. It’s grown from 13 or 14 sections a semester a dozen years ago to about 35 in fall and spring and 15 in the summer.

For masterminding this transformation, in 2012 the University awarded mathematics educator Clarkson and statistician Williams the third annual Chancellor’s Award for Excellence in Undergraduate Mathematics Instruction. The Office of Academic Affairs administers the award on behalf of Chancellor Goldstein, who earned a Ph.D. in statistics. The awards panel unanimously said these members of Hunter’s Department of Mathematics and Statistics stood out among the nominees for their long-term collaborative effort, innovation and emphasis on evidence-based instruction.

Today, the use of statistics is exploding in fields as diverse as nursing, psychology, sociology and political science, not to mention the media. “We, as citizens, are asked to make decisions based on them,” Clarkson says. Think of the last presidential election, with its dazzling graphics and dizzying polls. (This is a sweet spot for Williams, who formerly headed operations at the Louis Harris Associates polling firm.)

The professors began revising Statistics 113 by team teaching. Clarkson put students in groups, but they needed a spark.

In 1999, they found an early release of ActivStats multimedia software, written by Paul Velleman and distributed by Pearson. “There were videos on the disk. Students could read about some technique and press a button, start the [included] DataDesk software and analyze data using that technique,” Williams says.

“Students need to be active learners,” Clarkson says, and even the most brilliant lecturer has a hard time — especially these days, when students may text or do unrelated homework on their laptops on class time. Besides, she adds, students assume that “learning has to wait to take place until they go home and review their class notes.”

She recalled her days running Hunter’s remedial math program. She would look over students’ shoulders, correcting mistakes and explaining as they worked. “Students learned material more thoroughly and were more prepared for tests and didn’t feel we had any wasted time.”

So around 2003 or 2004, they broke with routine and shuttled students between a lecture hall and computer classrooms. “We wanted to make them more active participants and give them help when they ran into problems,” she says.

A year later, the professors did something radical. “Students did not want to give up their textbook, so we took it away from them,” Williams says.
With college funding, Clarkson and Williams stepped back from teaching the course to become orchestra conductors. They trained a cadre of part-time instructors for what by then were 700 students to be classroom helpers, rather than lecturers. At least 60 percent of today’s statistics instructors have been at Hunter for three or four years, while some have taught there 10 years or more.

“We have instructors showing students how to work software and do data analysis,” Williams says. “We started early on emphasizing the graphical side, for a picture is worth a thousand words,” both for understanding and conveying information.

The typical Statistics 113 student has completed a college algebra class or the equivalent, and English 120, the first writing course. Before class, students usually review a chapter in the software, looking at videos or models.

This year, each student has an individual, semester-long project based on a common set of some 6,000 data points. “They look at a variable and see if the data is skewed or normal and what the summary statistics are,” Clarkson says. After completing ongoing assignments, “at the end of the semester they write recommendations geared to someone who does not know statistics. They have to take something very technical and communicate it in a nontechnical way.”

The professors never stop refining the course, most recently by putting homework and the midterm exam online with the help of Pearson’s software and technical team. Since 2006, they’ve also used a uniform final for all sections that examines success in 10 learning outcomes. This way instructors don’t have to create or mark exams, freeing all classroom time for instruction.

So do students learn better this way?

The pass-rate percentages “are in the upper 80s,” Clarkson says. “The students who don’t pass are those who don’t finish the course, but there are not that many of them.”