Advancing Nursing Research with a Statistical Focus

Statistics is at the core of key figures in this world – unemployment rates, population growth rates and housing, school and medical facility needs. And Duke University School of Nursing is now making it an essential part of its Center for Nursing Research with the Research Design and Statistics Core.

Leading this core area is Wei Pan, PhD, associate professor and biostatistician. Pan’s research focuses on propensity score analysis, resampling, power and effect size, advanced statistical modeling, meta-analysis and their applications in the social, behavioral and health sciences. Pan recently edited a book entitled "Propensity Score Analysis: Fundamentals and Developments" that tackles both methodological and practical issues in propensity score analysis.

Before the new core was established, nurse researchers collaborated with statisticians from other academic fields such as sociology or psychology. However, School of Nursing leaders saw a need to provide nurse researchers with statisticians who could take the science of statistics and make it more applicable to nursing research.

According to Pan, nurse researchers typically use a simple or traditional statistical method for their research, even though other options are available. “There are more advanced methods that are being used by other fields that would allow greater scores during evaluation of research,” he said. “This in turn would increase the merit of grants for nurse researchers. This is something nursing research is just getting into while other areas such as medicine or psychology have been using advanced methods for a while. Nursing is getting there, it just takes time.”

Another disadvantage of using simple or traditional statistical methods is the possibility of not catching missing data – something that could be caught with more advanced methods such as multi-level modeling where data for participants are organized at more than one level.

With so many advanced methods available, the Center for Nursing Research wants to ensure that the School’s nurse researchers have a central location that allows them access to experts in research methods and receive more focused support specific to their line of work.

The Research Design and Statistics Core is staffed by five statisticians – three faculty statisticians, including Pan, and two master's-level statisticians. “Our new core provides nurse researchers with statisticians who are experts in a variety of methods such as propensity score analysis, sequential multiple assignment randomized trials or mediation and moderation,” said Pan. “We’re able to apply these methods to help advance nursing science.”

Part of the new core is a statistics lab that is open to faculty, students and postdocs two days a week. The lab provides research design and statistical guidance and consultation for research, including questionnaire design, measurement validation, data entry, data management, data analysis and statistical software coding. Besides onsite master's-level statisticians, the lab also offers the service of an on-call faculty statistician.

“It’s very rare that nursing schools have in-house statisticians that are accessible to faculty and students,” said Pan. “Providing in-house statisticians allows faculty and students to have meaningful interactions versus working with an adjunct professor from another field who may not understand the common language of nursing research. In-house statisticians can even help students become more comfortable in the research process, as well as provide more efficient communication.”

The benefit of an in-house statistics core isn’t just for students. Faculty are able to easily obtain support for pre-award grant applications and funded or unfunded projects. Statisticians are likewise able to use the experience to further develop their careers as they complete methodology work and teach research methods to doctor of nursing practice (DNP) and doctor of philosophy (PhD) students.

A United States Department of Labor report shows that the statistician occupation will grow 34 percent from 2014 to 2024 as a result of more widespread use of statistical analysis to make informed business and health care decisions. Pan agrees with these results but says there is a greater need. “We need more statisticians to develop methodology that is more applicable for nursing research,” he said. “There’s a need for more innovative methods to match the innovative work that is happening in nursing research.”

Scroll back to top automatically