Professor Jim Friedrich is a social psychologist at heart, although many students and faculty across campus know him as the resident statistician, the man they turn to when they need help quantifying their research. "I get to be the department dentist -- everybody comes to see me, but no one really wants to be there," he quips.
Discovering relationships among research observations and determining whether results are due to chance or something more -- these are the areas where statistics come into play. Since Friedrich joined Willamette's psychology department 15 years ago, he has used statistics to study everything from how people judge social science research to their attitudes toward affirmative action in college admissions. "I do a lot of research about how people form attitudes, how attitudes are changed and how people make decisions," he says.
College admission policies are ripe with statistics, and Friedrich and his students have examined numerous aspects of how undergraduates are accepted and whether they will succeed. One study led by a student tried to quantify whether the SAT exam was a solid predictor of success in college. The student collected data on an entire graduating Willamette class, including the students' SAT scores, their high school grades and their final college GPA. Results showed that both high school grades and SAT scores were fairly strong predictors of college GPA -- and that they were better predictors when used together rather than alone.
One of Friedrich's more recent interests involves what he calls the "naturalistic fallacy." When social scientists publish their research findings, the general public typically does not read those findings directly from the source. Instead, much of the research is reported to the public through newspaper articles or editorials, which may have an agenda or seek to interpret the findings, Friedrich says. "The science is simply describing nature. It's not making a value judgment on whether nature is good or bad. But people naturally draw moral inferences that are not there."
For example, if a statistical study suggests that men might be hard-wired to want a larger number of sexual partners than women, people reading about the report in the popular media might interpret the study as saying it's OK for men to have multiple partners, whereas women should stay monogamous. The study did not make this inference -- it simply reported on a statistical finding.
These types of interpretations also can affect public policy. A 1998 report on child abuse published by the American Psychological Association was officially condemned by Congress after the media, the public and other scientists were outraged by what they interpreted as immoral judgments in the report. "The report wasn't making any moral recommendations, but without even reading the actual article, Congress passed a measure condemning it," Friedrich says.
These types of issues raise a tough question for scientists: What is their ethical obligation when their published work could be misused or misinterpreted, especially if that work influences public policy?
"Nowadays you can hardly go on websites like Yahoo! News without seeing something about a researcher's findings on the brain, as an example," Friedrich says. "It's tough to guess how people will interpret and process this social science information. As researchers, we want to believe we have something to say, and we want to contribute important information to society. But we also have to understand how it might be transformed as it's communicated through non-professional outlets."