Researchers to Develop New Way to Analyze Study Data

April 12, 2006

WINSTON-SALEM, N.C. – A statistician and an economist at Wake Forest University School of Medicine are among 11 groups in the country awarded grants by the National Science Foundation to find new and better ways to use mathematical models to analyze data from social and behavioral science studies.

With a $350,000 grant, the Wake Forest researchers will develop and refine a method for analyzing data collected over time, allowing scientists to identify patterns of behavior in study participants that may help predict future behavior. This information can be used to improve treatments and to identify groups of individuals who may be at higher risk of failing to follow treatment recommendations.

The method has already been used in a study of people with schizophrenia that found that different patterns of side effects were associated with different treatments. In a separate study of people with malignant brain tumors, the researchers analyzed the patients’ emotional, cognitive, and physical symptoms. The analysis enabled them to identify a group of especially optimistic patients who showed relatively little emotional disturbance. Physicians can use this new information to more accurately decide which patients would benefit from mental health treatment.

“The new method can be applied to across a range of problems – from disability to mental health,” said Edward Ip, Ph.D., associate professor of biostatistical sciences and principal investigator of the project.

The goal of the NSF grants is to develop new research methods in social and behavioral sciences and economics. According to the NSF, advances in collecting data have created the need for new analysis models.

The new method – which the scientists call Temporal Configuration Analysis –builds on previous work by the research team and complements existing statistical methods. It will be developed over the next three years.

The basic idea behind the method is to help scientists make sense of complex data collected in studies that follow people over time. For example, in the study of brain tumor patients, all 100 patients in the study were rated on 13 different items related to their cognitive, emotional, and physical symptoms at five different points in time.

“Thinking about these multiple dimensions and variables is hard, because few of us can visualize more than two or three variables at a time,” said Alison Snow Jones, Ph.D., a co-principal investigator of the project, who is an economist and assistant professor of social sciences and health policy. “The method helps to group participants so scientists can see complex patterns of behaviors or symptoms over time.”

The scientists plan to apply the model to a variety of existing data, including an analysis of the impact of maternal alcoholism on children’s behavior problems. Another project is a study of North Carolina tobacco farmers to identify patterns in their attitudes toward risk and the impact of these patterns on efforts to diversify their crops.

Once the new method is developed and refined, it will be shared with other researchers through publications, at meetings, and through computer software. The researchers will also mentor high school, college and graduate students. For example, high school students who participate in a summer science and math enrichment program sponsored by the Center of Excellence for Research, Teaching and Learning at Wake Forest University will identify a simple research problem and apply the new method.

Working with the Wake Forest researchers as a co-principal investigator on the method is Steve Scott, Ph.D., from the University of Southern California.


Media Contacts: Karen Richardson,, or Shannon Koontz,, 336-716-4587.

Wake Forest University Baptist Medical Center is an academic health system comprised of North Carolina Baptist Hospital and Wake Forest University Health Sciences, which operates the university’s School of Medicine. The system comprises 1,187 acute care, psychiatric, rehabilitation and long-term care beds and is consistently ranked as one of “America’s Best Hospitals” by U.S. News & World Report.

Media Relations