Researchers Benefit from Center’s Statistics Experts
A theory not backed up by cold, hard facts is just that—a theory.
But in the realm of social work, securing strong evidence to reinforce an assumption is not always an easy task. Variables are numerous and often difficult to measure. Investigators typically examine specific groups rather than the general populace, further complicating the process.
However, researchers at the USC School of Social Work’s Hamovitch Center for Science in the Human Services have an edge when it comes to the often perplexing world of data: a team of experts in statistical design, data management and analysis that serves to ensure studies are robust and empirically sound.
And although investigators have benefited from biostatistical assistance provided by center staff for years, a recent effort to strengthen and expand the statistics team is significantly bolstering the level of support.
“Now it’s a more coherent effort,” said Chih-Ping Chou, the center’s senior biostatistician who holds a joint appointment as a professor of preventive medicine in the Keck School of Medicine of USC. “We can help not just on the grant writing and statistical analysis, but we can also provide more comprehensive data management.”
The re-imagined Data Analysis and Management Core is designed to address a wide array of research needs at the center, ranging from basic technical assistance with grant proposals to consultation on advanced statistical techniques.
In addition to assisting faculty researchers as they envision and develop their study designs, the team helps validate new measurement scales and tools created by investigators, strengthening the value of study findings. Once data is gathered, core members offer support with analysis and manuscript preparation, as well as the design and management of databases. Finally, the team is seeking to enhance multidisciplinary skills by training researchers and students in statistical methodology.
“We’re developing our own entity as a core, almost like a mini-department of social statistics,” said Charles Kaplan, the center’s associate dean of research who is responsible for overseeing the statistics group.
The effort to increase statistical support at the center can be traced back several years to when school leaders began encouraging faculty members to submit study proposals. A key component of crafting a successful grant proposal is outlining a clear and concise plan for data collection and analysis, Chou said.
As the head of statistical analysis and measurement cores for multiple federally funded transdisciplinary research centers based at USC, Chou has a strong background in study design and analytical approaches with regard to grant writing. He described his primary role in the Hamovitch Center’s statistics core as providing assistance to researchers at the conceptual level—as they prepare their study proposals and design their data analysis plans.
Conversely, as the center’s biostatistician, Lei Duan is more involved in the day-to-day process of analyzing data and outlining specific statistical methods for research proposals and journal manuscripts. Duan studied under Chou as a doctoral student and worked as a statistician on cancer research studies before joining the Hamovitch Center.
“I enjoy the challenge here,” she said, explaining that studies involving psychological or behavioral health data utilize a wider array of models and skills than epidemiological research.
In social work, researchers often work with quasi-experimental data and complex models in which multiple variables must be considered simultaneously to ensure the correct relationship is being examined.
“You have to do a lot of validation to make sure everything is correct,” she said, adding, “I want to use the most advanced techniques to solve the problem, to solve the research questions.”
As the team’s new data manager, Pey-Jiuan Lee, is largely responsible for organizing and managing information once it has been gathered. In recent years, she has provided database support for several specific grant-funded projects.
However, as the statistics core expands its data management resources, Lee is looking forward to meeting with faculty members and determining what services are in demand.
“We want to see what kind of [data management] services our school researchers need the most,” she said.
In essence, the team plans to develop a web-based data warehouse to expedite the data entry process, Kaplan said, in addition to compiling and sharing information across research projects. The new system also would address security and reliability issues. “Before now, it’s largely been left up to the individual researchers to back up their data,” he said.
As the statistics team begins to take on more responsibilities, core leaders acknowledged that developing training seminars on statistical software and methodologies will be critical to the success of the venture.
In addition to offering training sessions to investigators and staff involved with research projects, Chou said the team plans to reach out to students interested in research and social statistics.
“I feel the PhD students will provide a very good resource, especially if they are working as research assistants,” he said.
And although the statistics core is still in the process of building up its service capacity, Chou said the team is excited to begin working with researchers on new projects, as well as developing a better understanding of what resources will be needed for the future.
“I think this will be a very iterative process,” he said, “and I also hope that through this process, we can also involve students even more.”
- Master of Social Work