Post-doctoral Positions: A Path from Graduate School to Career
Jamie Nunnelly, Communications Director for National Institute of Statistical Sciences
Making the transition between graduate school and a new career can be daunting. Academia guides people each step of the way, checking a person’s work and telling them what to do and how to think about a particular topic, but life outside of the classroom is different.
One way to ease this transition is to take a post-doctoral position. These positions last for one to three years and allow a person entering the work force an opportunity to think and act independently.
The National Institute of Statistical Sciences (NISS) offers a post-doctoral program that is one of the largest, and most successful in statistics. It has provided strong impetus to more than 80 early career researchers, who have all gone on to successful careers.
One of the post-doctoral fellows who benefited from the NISS program is George Luta, an assistant professor in the department of biostatistics, bioinformatics, and biomathematics at Georgetown University Medical Center. The department is part of the Lombardi Comprehensive Cancer Center (LCCC) at Georgetown, so much of Luta’s activity involves collaborative cancer research.
Luta studied at The University of North Carolina at Chapel Hill (UNC). He heard NISS offered post-doctoral positions as he was finishing his PhD in biostatistics and decided to apply. He was a post-doctoral fellow at NISS from August 2006 to August 2007 and undertook varied assignments. One was with Alan Karr, director of NISS, on an expert task force on effect sizes (which distinguish “practical significance” from statistical significance in a domain-specific manner) for the National Center for Education Statistics (NCES). Luta undertook an extensive background study on effect sizes, which enabled the task force to make specific actionable recommendations to NCES.
Luta also worked with Nell Sedransk, associate director of NISS, on a research project involving the measurement of the QT intervals for electrocardiograms. The length of the QT interval, which can be indicative of adverse side effects, is often used when evaluating a new drug going through the U.S. Food and Drug Administration’s approval process. NISS worked on developing new methodology to measure the QT intervals. Luta and Sedransk identified the scope of work and started getting the right cross-disciplinary partners to join in the research. Later, S. Stanley Young, assistant director for bioinformatics at NISS, and some of his former colleagues from Eli Lilly also worked on a QT interval project. The research led to a paper by Luta and the colleagues at Lilly, “Sample Size Calculations in Thorough QT Studies,” which appeared in the Journal of Biopharmaceutical Statistics in 2008.
“I really enjoyed being part of an interdisciplinary team,” remarked Luta about his experience at NISS. “The teams would write white papers describing step by step what needs to be done to solve a specific problem, and then they will start doing the research work. What I do now at Georgetown is a continuation of that type of collaborative research. It’s what they call team science or ensemble science. … Team-based research makes the statistician an equal player at the table and recognizes intellectual contributions above and beyond his/her statistical expertise,” noted Luta.
Collaborating with other post-docs was also something Luta enjoyed during his time at NISS. He and Michael Last, another NISS post-doc, collaborated on a research project also involving Young. This resulted in a paper, “Pooled ANOVA,” that was published in 2008 in Computational Statistics and Data Analysis.
As Luta’s year was ending at NISS, the staff helped him prepare for job interviews. They gave him feedback on his presentations, helped him prepare relevant questions to ask, and gave him advice about the interviewing process. Luta interviewed with Georgetown University while he was still at NISS.
Currently, Luta is collaborating with researchers from the Lombardi Comprehensive Cancer Center on cancer control and prevention, although some of his research work is in bioinformatics. Luta said he learned bioinformatics methods while at UNC and NISS. Specifically, Young introduced him to the non-negative matrix factorization, a methodology that he used while participating with two of his students in a competition to analyze flow cytometry data. The results from the competition were published in Nature Methods this year in a paper titled, “Critical Assessment of Automated Flow Cytometry Data Analysis Techniques.” Luta also contributed at a session organized by Young at JSM 2013 on non-negative matrix factorization.
Luta pointed out that it was really great that the Statistical and Applied Mathematical Sciences Institute (SAMSI) is located in the same building as NISS, allowing NISS post-docs the opportunity to attend SAMSI workshops and interact with SAMSI post-docs.
NISS typically advertises for post-doctoral positions as new research projects get under way. Check the NISS website to find out about new positions.