No

# Dr. Stats on call

### June 12, 2008

Document Tools

Font Size
S      M      L      XL

When Tom Loughin was eight years old he would pour out a bowl of Lucky Charms cereal and organize the different coloured marshmallows into a histogram to see which colour had the largest number. "But I didn’t keep records at the time," the SFU Surrey statistician confesses. "I was too impatient, I just ate them."

Today, Loughin operates the statistical consulting service at SFU’s Surrey campus. The service started at the Burnaby campus in 1980, advising students and faculty on statistical problems. Now, there is an online version, Dr. Stats. (SFU faculty and students can submit their statistical problems to the service using an online form at www.stat.sfu.ca/people/consulting/drstats).

"I’m like the 1970s television detective, Columbo," says Loughin. "I keep asking stupid-sounding questions until I have a complete picture and can help you identify what you need to do." The most common problem, he says, is a haphazard study design. "It’s very important that you identify the population of things you’re trying to draw a conclusion about, and that the data that you gather is representative of that population."

A famous example of a study based on the wrong population, he says, was a 1948 newspaper headline saying that Thomas Dewey had won the American presidential election. The newspaper based its story on a telephone poll, but at that time telephones tended to be owned by upper or upper-middle class people — the group most likely to be supporting Dewey — so they were over-sampled in the poll. "Have you ever heard of President Dewey?" Loughin asks. "No, because Harry Truman won that election."

Loughin came to SFU Surrey from Kansas State University, where he filled a similar role, helping agricultural researchers design, analyze and interpret studies. For instance, he helped them develop new varieties of wheat. "From seed to stomach, I’ve studied wheat," he says. "I don’t know anything about it though, except statistically."

Today Loughin’s role is basically the same. "Statisticians don’t grow the wheat. And statisticians don’t do perceptual research. But we make that research better."

In spring of this year, Loughin was named a fellow of the American Statistical Association for his exemplary research in categorical data analysis; outstanding teaching, consulting and editorial contributions; effective promotion of statistics to a wide array of audiences, and for leadership in service to the profession.
Search SFU News Online